Sunteți pe pagina 1din 62

The Institution of Structural Engineers

The Department for Communities and Local Government


The Department for Culture Media and Sport

December 2008

Dynamic performance
requirements for permanent
grandstands subject to
crowd action

Recommendations for management,


design and assessment

The Institution of Structural Engineers


International HQ, 11 Upper Belgrave Street, London SW1X 8BH, United Kingdom
T: +44 (0) 20 7235 4535
F: +44 (0) 20 7235 4294
E: mail@istructe.org
W: www.istructe.org Published by the Institution of Structural Engineers

Stadia covers.indd 2-3 3/12/08 16:13:11


The Institution of Structural Engineers
The Department for Communities and Local Government
The Department for Culture Media and Sport

December 2008

Dynamic performance
requirements for permanent
grandstands subject to
crowd action

Recommendations for management,


design and assessment

Published by the Institution of Structural Engineers


Membership of the Joint Working Group

Dr J W Dougill – Chairman
Professor A Blakeborough – Oxford University
Mr P Cooper – KW Ltd to July 2007, then INTEC
Dr S M Doran – IStructE, Secretary
Dr B Ellis – BRE to 2006 now Consultant
Mr P F Everall – DCLG (to 2005)
Dr T Ji – UMIST/ The University of Manchester
Mr J Levison – Football Licensing Authority, d. 12th Dec. 2007
Dr J Maguire – Lloyd’s Register. (received papers from 2002)
Mr S Morley – Bianchi Morley
Mr M Otlet – W S Atkins
Professor G A R Parke – Surrey University (to 2002)
Mr J. G. Parkhouse – Parkhouse Consultants
Professor A Pavic – The University of Sheffield
Mr L Railton – Health and Safety Executive (to 2003)
Mr W Reid – Consultant, URS
Mr R Shipman, DCLG, (from 2005)
Mr P Westbury – Buro Happold. (received papers from 2002)
Mr M Willford – Ove Arup
Professor J Wright –The University of Manchester/J2W Consulting Ltd

Corresponding Members
Mr D Allen, National Research Council of Canada
Dr M Kasperski, Bochum University
Dr P Reynolds, The University of Sheffield
Mr P Wright, Health and Safety Executive, from 2003

Contributors
The Joint Working Group wishes to acknowledge the contribution of the following individuals who, though
not attending as members of the Group, made presentations on different aspects relating to grandstand
design, operation and behaviour.

Mr D Allen, National Research Council of Canada


Mr J Cutlack, J. Bobrowski and Partners
Dr J Dickie, Crowdsafe Ltd
Mr C Gleeson, Chelsea Football Club
Professor M J Griffin, Institute of Sound and Vibration, Southampton University
Dr M Kasperski, Bochum University, Germany
Dr J Littler, BRE
Dr A J Soane, Bingham Cotterell

Published by the Institution of Structural Engineers


International HQ, 11 Upper Belgrave Street, London SW1X 8BH, UK

ISBN: 978-1-906335-12-0

© 2008: The Institution of Structural Engineers

The Institution of Structural Engineers, DCLG, DCMS and the members who served on the Joint Working Group
which produced this report have endeavoured to ensure the accuracy of its contents. However, the guidance and
recommendations given in the report should always be reviewed by those using it in the light of the facts of their
particular case and specialist advice obtained as necessary. No liability for negligence or otherwise in relation to this
report and its contents is accepted by the Institution, the members of the Joint Working Group their servants or agents.
Any person using this report should pay particular attention to the provisions of this Condition.
No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means
without prior permission of the Institution of Structural Engineers, who may be contacted at 11 Upper Belgrave Street,
London SW1X 8BH.

ii Dynamic performance requirements for permanent grandstands subject to crowd action


Contents

Foreword v
1 Scope of the Recommendations 1
2 Design event scenarios 2
3 Listed Engineers  4
3.1 Requirement for specialist engineering expertise 4
3.2 Technical support  4
3.3 Discretion on relevance of the Recommendations to specific structures 4
3.4 Judgment on relevance of the Recommendations to specific structures 4
3.5 Monitoring 5

4 Natural frequencies and other dynamic properties 6


4.1 Background 6
4.2 Structural modelling to determine modal properties of the
empty grandstand 6
4.3 Values for initial design 7
4.4 ‘Relevant’ natural frequency  7

5 Testing 8
5.1 Need for testing 8
5.2 Aims of testing  8
5.3 Circumstances requiring testing  8

6 Management responsibilities  10
6.1 Overall responsibility 10
6.2 Design of new stands  10
6.3 Change of use and assessment for specific events  10
6.4 Operational strategies to reduce dynamic response and crowd alarm 11
6.5 Handover of new and structurally modified grandstands 12
6.6 Operations Manual 12
6.7 Operation  13

7 Route 1: Compliance with natural frequency requirements 15


7.1 Vertical excitation 15
7.2 Side-to-side horizontal excitation 15
7.3 Nodding modes due to front-to-back excitation 15

8 Route 2: Design for managed events 17


8.1 Outline 17
8.2 Idealised description of crowd activity 17
8.3 Serviceability: Tolerance of motion 18
8.4 Serviceability: Displacement limits 18
8.5 Ultimate load capacity 18
8.6 Fatigue 18

9 Analysis of dynamic performance  19


9.1 Impulse loads  19
9.2 Horizontal loads due to periodic excitation 19
9.3 Analysis for vertical periodic excitation 19
9.4 Human structure interaction 19

10 Use of the Recommendations 21

Dynamic performance requirements for permanent grandstands subject to crowd action iii
Appendix 1  Background to human structure interaction 22
A1.1 Introduction 22
A1.2 Basic principles  22
A1.2.1 Modelling human structure interaction 22
A1.2.2 Active and passive behaviour 23
A1.3 Application of the theory 23
A1.3.1 Direct application of the theory 23
A1.3.2 Approximate analysis using an assumed mode shape for
the crowd’s motion 24
A1.4 Body Unit properties and loadings 24
A1.5 Analysis and results  24
A1.5.1 Modal analysis 24
A1.5.2 Root mean square (RMS) accelerations and acceleration limits 24
A1.5.3 Analysis with a dominant mode. 25
A1.5.4 Multi-mode analysis 25
A1.6 References 26

Appendix 2  Body unit properties and recommended loading 27


A2.1 Body unit and structure 27
A2.2 Crowd body elements 27
A2.3 Representation of periodic loading 28
A2.4 Internal ‘drivers’- Gi – producing dynamic crowd loading 29
A2.5 The crowd effectiveness factor ‘t’ 30
A2.5.1 Scenario 4 30
A2.5.2 Scenarios 2 and 3 30
A2.6 Monitoring and back analysis  31
A2.7 References 32

Appendix 3  Calculation of modal properties  33


A3.1 Introduction 33
A3.2 Modal analysis and natural frequencies 33
A3.3 Basic errors: two prime suspects 35
A3.3.1 Distinction between force and mass 35
A3.3.2 Stiffness 35
A.3.4 Methods for calculation of natural frequencies 36
A3.4.1 General comment 36
A3.4.2 Approximate analysis 36
A3.4.3 Computer based analysis 36
A3.5 Consequences of mistaken idealisations 36
A3.6 Comment 37
A3.7 Further information 38

Appendix 4  Dynamic testing of grandstands and seating decks 39


A4.1 Introduction 39
A4.2 What should be tested and what results are needed? 39
A4.3 Analysis and testing 40
A4.4 Principles of dynamic testing 41
A4.5 Excitation sources and testing techniques 43
A.4.5.1 Ambient vibration survey (AVS) 43
A.4.5.2 Heel-drop testing 43
A.4.5.3 Measured impact testing 43
A4.5.4 Shaker testing of different types and complexity 44
A.4.5.5 Future developments 45
A4.6 Specification and procurement 45
A4.7 Reporting 47
A4.8 Further Information 47

Appendix 5  Bibliography 48

iv Dynamic performance requirements for permanent grandstands subject to crowd action


FOREWORD

The Joint Working Group met first in January 2000. Its Interim Guidance was published in
November 2002 in response to concerns over crowd action on structures generally and on the
relevance of available recommendations to dense crowd loading on permanent grandstands. The
Interim Guidance used the vertical natural frequency, for the mode of vibration that could be
excited and felt by people on the seating deck, as the currency to determine different categories
of permissible use. No attempt was made to recommend a method of estimating performance by
calculation as it was considered that existing procedures, though widely used, could not be relied
on. The Interim Guidance was designed to be safe and straightforward to apply. It provided a
significant relaxation of the ‘trigger value’ frequency limits of BS6399 (1996) and the 1997 Green
Guide. However, because of the simplification of using natural frequency as the single factor
determining a category of use, it was a broad brush treatment. What was needed was a method
of design and operation that was based on an estimate of performance that was reasonable when
compared with the effects observed with active crowds in real structures. This was the task of the
Joint Working Group from 2002 onwards.
In addressing the technical issues relating to the analysis of the structure, the Joint Working
Group has been closely involved with a number of UK research projects (almost all supported
by the Engineering and Physical Sciences Research Council, EPSRC) that have been undertaken
since 2000 and which have contributed to an improved understanding of the physical problem of
how human beings interact with moderately flexible structures. This work has provided the basis
for the technical content in the new Recommendations. However, this could not be seen as an end
in itself. A key issue in design is how to deal with uncertainty. With grandstands, there can be no
absolute certainty on the way any random group of people will behave. Accordingly, the technical
provisions have been set in a framework of procurement, management and operation aimed at
minimising risk by managing uncertainty.
The Recommendations propose that specially ‘Listed Engineers’, having particular
experience and capability, be used in the design and assessment of grandstands for dynamic
crowd loading. Design should be based on the concept of managed events described by standard
Design Event Scenarios that form part of the specification for a stand. Within these scenarios, the
Management of the facility takes responsibility for specific agreed measures to mitigate the effects
of motion. Hand-over procedures are outlined with the aim of ensuring that the design calculations
relate to the as-built structure. The aim of each of these recommendations is to reduce uncertainty
where it is possible to do so.
The Recommendations are written for everyone who has responsibility for grandstands. This
includes the owners, operators, managers, architects, insurers and engineering designers as well as
Local Authority staff dealing with building control and safety issues. The Recommendations are
accompanied by Appendices directed particularly at the engineering analyst and designer.
The Recommendations relating to specification, management and operation should be
considered to be of equal importance to providing comfort and safety as the technical guidance
addressed principally to the engineering designer.

Dr John W Dougill
December 2007

Dynamic performance requirements for permanent grandstands subject to crowd action v


vi Dynamic performance requirements for permanent grandstands subject to crowd action
1 Scope of the Recommendations

The Recommendations given here are for use in the design or assessment of permanent
grandstands and relate solely to dynamic action due to crowd activity. Reference should
be made to other guidance and relevant Standards for other load cases and design
requirements.
The Recommendations also include guidance relating to Management’s role in
design and in implementing operational requirements for stands being used for events
at which dynamic crowd loading can be expected. This guidance is concerned only with
those aspects of crowd management that directly influence the structural response of a
grandstand and so supplement, but do not replace legal and Standards’ based requirements
for safe operation. The operational arrangements adopted should be taken into account by
those undertaking an overall risk assessment at the scheme design stage or for a specific
event.
The Recommendations apply to grandstands with seating decks constructed in
structural steel, reinforced or prestressed concrete and combinations of these forms of
construction. No recommendations are made for the use of subsidiary systems to provide
additional damping or active control.
The Recommendations are considered relevant to grandstands with seating decks
having a supported span greater than 6m or cantilever spans of more than 2.5m. However,
it is recognised that, even within the declared scope of these Recommendations, there may
be particular grandstands for which the layout, form of construction or limited use might
render it unnecessary to undertake a full check on dynamic performance. The manner in
which this can be dealt with is treated in Section 3 on Listed Engineers.
The Recommendations revise and extend the recommendations given in the November
2001 report ‘Dynamic performance requirements for permanent grandstands subject to
crowd action: interim guidance on assessment and design’, published by the Institution of
Structural Engineers and adopted by DCLG and DCMS. This Interim Guidance used the
vertical natural frequency of the empty grandstand as the sole criterion for assessing the
acceptability of grandstands for use with crowds likely to generate dynamic loading. This
convenient, but coarse grained, approach is retained as an option in the present guidance
which now provides an alternative approach that depends on engineering estimates of the
likely performance of a grandstand for events at which the event organiser is responsible
for specific agreed measures relating to crowd management. This alternative approach
provides further options for the designer and management as well as addressing a need to
account for influences on behaviour additional to natural frequency.
As a consequence of this approach, there are two Routes available for design and
assessment of grandstand structures subject to dynamic crowd loading,
• Route 1: Based on limiting values of natural frequency for the grandstand empty of
people.
• Route 2: Based on estimates of performance of grandstands calculated for specified
managed events.

A grandstand may be considered to meet the Recommendations for dynamic crowd loading
if the requirements of either one or other of these routes are met together with the separate
conditions for horizontal strength and stability. (See Sections 7.2, 7.3 and Table 2).

Dynamic performance requirements for permanent grandstands subject to crowd action 1


2 Design Event Scenarios

The Design Event Scenario forms the basis for a design specification for dynamic
performance of a grandstand under crowd loading. Table 1 gives standard Design
Event Scenarios for use in communicating design objectives in terms of anticipated
performance.
The Table shows a range of different events together with the expected activity of
the crowd and an indication of crowd control measures to be provided by the management
of the grandstand.
The four performance based scenarios (numbered 1 to 4) correspond to increasing
crowd involvement and activity together with increased loading. Scenarios 1 and 2,
appropriate for viewing sporting events and classical concerts, would normally be satisfied
by Route 1 requirements. Scenario 3 refers to lively concerts and high profile sporting
events whilst Scenario 4 is for high energy events such as pop/rock concerts. The Scenarios
are provided in order to assist event specific assessment and to provide a yardstick for
authorities concerned with safety certification.
It will be evident that the Design Event Scenario is the statement of what should
be covered in design and what needs to be managed. The Scenario comprises a reference
to the category of event in Table 1 with a statement of any additional specific crowd
control measures that have been agreed as being required. (See Section 6.4 on Operational
Strategies).
It should be noted that the descriptions of exemplar events are indicative rather
than prescriptive. For example, an event may be described as a pop-concert for publicity
purposes but the crowd’s reaction may be only moderate and so more consistent with a
concert with medium tempo music as envisaged for Scenario 3. Accordingly, in using
records of past events to assist an assessment for a future event, care should be taken to do
this on the basis of observed performance and not solely on a record that a ‘pop-concert’
had been run satisfactorily in the past. (See Section 6.7 on Operation, re. record keeping).
The Scenarios are based on experience of events in the United Kingdom. In assessing
any specific event, judgment will be needed to decide the appropriate category particularly
between Scenarios 3 and 4 and, on occasion, whether the crowd at a particular sporting
event is likely to be more than usually active with coordinated rhythmic activity. This
has become common at football matches in mainland Europe where groups of fans have
rehearsed bobbing, treading or stamping in time to a beat provided by their leader. The
improved coordination accompanying behaviour of this sort can lead to motion that is more
severe than that anticipated for Scenario 4. This could lead to possible discomfort for seated
or standing fans not participating in the activity. Such situations need to be recognised and
appropriate operational measures adopted by Management. (See Section 6.4 on Operational
strategies).

2 Dynamic performance requirements for permanent grandstands subject to crowd action


Table 1 Event Scenarios to be supplemented with a statement on agreed operational arrangements
Scenario Exemplar event Crowd behaviour Management Acceptance criteria
Expected crowd make up Stewarding Crowd expectation Natural frequency or
design acceleration
limit
1 Stand used for viewing of Normally relaxed Predominantly seated Stewards require to Comfort Route 1 with 3.5 Hz
sporting and similar events viewing public with be instructed as to min. or accepted at
with less than maximum spontaneous response the possible effects the discretion of a
attendance to single events of motion of the Listed Engineer
structure resulting
from coordinated
2 Classical concert and Audience seated with Predominantly seated Route 1 with
crowd activity and,
typical well attended only few exceptions – Comfort 3.5 Hz min. Is
particularly for
sporting event minor excitation normally adequate
Scenarios 3 and 4,
but otherwise
need to be provided
Route 2 and 3%g max.
with the means for
RMS acceleration
instant communication
with central control.
3 Commonly occurring Potentially excitable All standing and A few individuals may Route 1 with 6 Hz min.
events including, inter alia, crowd with crowd participating during some complain at lack of Otherwise Route 2,
high profile sporting events participation part of the programme comfort but most will and 7½%g max. RMS
and concerts with medium tolerate the motion acceleration
tempo music and revival
pop-concerts with cross
generation appeal

Dynamic performance requirements for permanent grandstands subject to crowd action


4 More extreme events Excited crowd, mostly Mainly young Excitement and Route 1 with 6 Hz min.
including high energy standing and bobbing and active with motion but with Otherwise Route 2,
concerts with periods of with some jumping vigorous participation expectation of and 20%g max. RMS
high intensity music personal safety acceleration
Note
Appendix 1, Section A1.5 deals with Root Mean Square (RMS) accelerations and how these are defined and used in analysis.

3
3  Listed Engineers

3.1 Requirement for specialist engineering expertise


Grandstands, and particularly the combination of a crowd of people and a grandstand,
provide technical and managerial problems that can be addressed satisfactorily only
by engineers with appropriate specialist expertise in addition to all round competence
underpinned by professional qualifications.
Whichever Route is chosen for design or assessment, it is important that these functions,
and the provision of advice on safe operation, are undertaken by an engineer having relevant
expertise in the structural design and safe operation of grandstands. At the request of
Government, the Institution of Structural Engineers is putting in place a scheme to provide
a list of engineers with specific experience in the design, assessment and safe operation of
grandstands for dynamic crowd loading. In these Recommendations, such engineers will be
referred to as Listed Engineers.
It is proposed that any Design Team dealing with new grandstands or significant
alteration to existing grandstands should include a Listed Engineer with particular
responsibility for overseeing those aspects of design concerned with dynamic crowd
behaviour. Similarly, a Listed Engineer should be employed in assessing existing stadia
and in overseeing the hand-over procedures of new and altered grandstands.

3.2 Technical support


Listed Engineers will be expected to have sufficient personal specialised knowledge and
expertise relating to the design and safe operation of grandstands to advise management
on all aspects connected with dynamic crowd loading. However, it would be expected that
Listed Engineers themselves may need to employ specialist support in respect of physical
testing and the conduct of dynamic structural analysis undertaken under their direction or
on their instruction.

3.3 Discretion on relevance of the Recommendations to specific structures


Grandstands vary in type, size and manner of use. Depending on the circumstances, detailed
consideration of crowd action may not be appropriate for a particular grandstand and,
based on their experience and technical background, Listed Engineers will be expected to
use their discretion in deciding the relevance of the Recommendations to such cases.
The Listed Engineer’s report to the grandstand’s Management should make clear
whether discretion has been exercised and the reason for doing so. It will be appreciated
that the management of a grandstand has responsibility for the safety of the facility
and so needs to understand and take responsibility for accepting the Listed Engineer’s
recommendations. The Listed Engineer’s report should also be made available to the Local
Authority concerned with Building Control and Safety Certification.

3.4 Judgment on relevance of the Recommendations to specific


structures
Listed Engineers will be expected to have the necessary background to use their own
judgment in interpreting and applying the Recommendations to particular grandstands.
The main areas for exercising judgment will normally be in advising the client
when choosing the appropriate Design Event Scenario, in detailed consideration of the
approximations involved in structural modelling and in assessing the need for structural
testing.

4 Dynamic performance requirements for permanent grandstands subject to crowd action


In making these judgments, the Listed Engineer should bear in mind that, in respect
of dynamic behaviour, the aim should be to remove the need to consider structural
behaviour from the wider risk assessment necessary for an event as a whole. If followed,
the Route 1 approach provides for this based on measured natural frequencies and the
frequency limits in Figure 1. (See Section 6.6). The Route 2 approach provides more
flexibility in decision making but depends on estimates of performance using calculations
based on recommended values for crowd loading that are considered typical for different
types of events. Sufficiently accurate values of natural frequency are required for both
approaches whilst a Route 2 analysis needs a more complete knowledge of dynamic
behaviour including a full set of relevant modal properties.
Because of the importance of accurate knowledge of dynamic properties, either
in characterising the admissible use of a grandstand using Route 1, or, as the basis of
design or assessment by Route 2, it will normally be considered necessary for the values of
dynamic properties to be established, or checked, using physical testing. (See Section 5 on
Testing). There will be circumstances where testing is not necessary, so placing a particular
responsibility on the Listed Engineer to exercise personal judgment on the need for physical
testing to check or establish values of dynamic properties. Such a judgment should not be
influenced by considerations of cost, convenience or time pressure but, solely, on a decision
that, in the particular circumstances being considered, the wider information available
from testing would not materially affect decisions on either design or assessment.
As in the use of discretion on the relevance of Recommendations to a particular
grandstand, the Listed Engineer’s report to the grandstand’s management should make
clear whether personal judgment has been exercised on any particular Recommendation,
including the need for testing, and the reason for doing so. The Listed Engineer’s report
should also be made available to the Local Authority concerned with Building Control and
Safety Certification.

3.5 Monitoring
The Listed Engineer should be involved with and advise on any programme to monitor the
behaviour of a stand whilst in use under crowd loading. Monitoring could include visual
recording of crowd behaviour, acceleration and load measurements, stewards’ reports etc.

Dynamic performance requirements for permanent grandstands subject to crowd action 5


4 Natural Frequencies and other dynamic properties

4.1 Background
Partly as a result of testing undertaken since the adoption of the Interim Guidance, the
concern on possible differences between calculated values of natural frequency and values
determined by competent testing has been reinforced rather than allayed. Differences of up
to 30% between measured and calculated natural frequencies have been recorded. Also, it
is rare that even approximate agreement between calculated values and those from testing
is obtained without some reappraisal of the structural model used in the calculation.
The largest differences in values usually follow from misguided initial qualitative
assessments of likely structural behaviour that then determine the form and extent of
the structural model used in the calculation. However, even with carefully considered
structural models and with the most diligent attention to detail, differences of up to 15%
between calculated and physically determined values are common. These differences can
be expected. The structural model will normally be based on assumed material properties
and idealisations of the connectivity between structural elements comprising the grandstand
structure together with assumptions on how much of the structure needs to be modelled. In
addition, there will be uncertainties in the contribution of mass and stiffness from the non-
structural elements. In contrast to this, the as-built structure responds to excitation, either
in a test environment or due to crowd loading, according to how it is actually constructed
and maintained.
The possible discrepancy between values of dynamic properties used in calculation and
those found by testing may be sufficient to affect an assessment of performance based on the
Route 2 method or a determination of the appropriate category of use by the Route 1 method.
The uncertainty attached to using values obtained by calculation alone can be minimised
by physical testing of the structure while empty of people; either as part of the hand-over
procedures for new or modified structures or as part of a subsequent assessment.

4.2 Structural modelling to determine modal properties of the empty


grandstand
Although any form of linear elastic dynamic analysis with linear damping may be used,
modal analysis using a finite element representation of the grandstand structure is likely
to be the preferred method except for all but the simplest structures. Appendix 5 provides
details of the theory underlying the method and notes on modelling grandstand structures
for calculation of natural frequencies are given in Appendix 3. These notes indicate
aspects where particular care needs to be taken in developing the finite element model so
that all factors influencing the dynamic characteristics of the grandstand are adequately
represented.
The structure should be analysed to determine the natural frequencies, mode
shapes and related modal masses in the absence of people, but including the mass of all
fixtures and fittings that are involved in motion of the structure. The modes that indicate
significant motion of the seating deck, and are capable of being excited by people on the
deck, need to be identified, together with their natural frequencies. The lowest of these
natural frequencies is taken to be the natural frequency for vertical excitation of the empty
grandstand to be used in the Route 1 design or assessment. (See Section 4.4 on Relevant
Natural Frequency). A Route 2 estimate of performance should be based on all the modes
that are considered to be capable of providing significant motion of the seating deck.

6 Dynamic performance requirements for permanent grandstands subject to crowd action


4.3 Values for initial design
Engineers undertaking the design of new structures should be aware of the potential
difference between natural frequencies calculated using even the best practice and the
values that will obtain in the completed structure. In design, it will be prudent to allow
for a minimum plus or minus 0.5 Hz variation between calculated values of natural
frequency and those that will influence performance in the actual structure. Variations of
this magnitude are not unusual in comparing calculated and measured values of natural
frequency and can have significant effect on estimates of performance.

4.4 ‘Relevant’ natural frequency


Design or assessment using the Route 1 approach is determined by a single value of natural
frequency. It is important to note that this is the relevant natural frequency that corresponds
to the mode of vibration with the lowest natural frequency at which people can excite the
seating deck and feel its motion.
It should be noted that the relevant natural frequency may not be the lowest natural
frequency that is determined by testing or analysis. Testing might show that a low frequency
mode exists but this needs to be assessed to determine whether a crowd on the seating deck
can excite the mode significantly.
As an example, consider a roof that vibrates with large amplitude vibrations at low
frequency when excited by wind action. If the roof and the seating deck are connected
so that there is a ‘vibration path’ between them, the effect of the ‘roof mode’ might be
found in tests on the seating deck even though the amplitude of the deck displacement is
small compared to that of the roof. In such cases, decisions on whether a particular mode
is, or is not, ‘relevant’ are better based if the values of natural frequency from testing are
supplemented by additional data on other modal properties. Also, inclusion of ambient
testing (See Appendix 4) in a test programme would provide information that could help to
identify modes driven primarily by sources separate from a crowd on the seating deck.
Testing may also reveal the occurrence of so called global modes of vibration in
which the whole grandstand, concourses, seating deck and roof may be involved in front-
to-back, sway or torsional motion. Again the ‘relevant’ natural frequency is the lowest
frequency corresponding to those modes by which people can excite the seating deck
through vertical movement and feel its motion.
In some global mode cases, such as with front–to-back ‘nodding’ modes of upper
cantilevers due to flexibility in the main supporting structure, it may not be possible to
decide, from the results of testing alone, which is the appropriate relevant natural frequency
on which to base an assessment or indeed whether it is possible that several modes might
be simultaneously excited by crowd action. In such cases, it will be necessary to use the
Route 2 approach and supplement the assessment with calculations of performance using
more than one mode of vibration.

Dynamic performance requirements for permanent grandstands subject to crowd action 7


5  Testing

5.1 Need for testing


In deciding on the need for testing, both the Listed Engineer and the Client should be
aware of the uncertainties in determining dynamic properties solely by calculation and
the benefits of obtaining confirmation of the values used in design or assessment through
physical testing of the as-built structure. The Listed Engineer should be fully involved in
advising on the choice of a competent Test Agency with appropriate expertise for testing
grandstands and should advise on the test programme to be adopted.
A revised version of the guidance on testing, first published by the Institution of
Structural Engineers in 2002 as the Advisory Note, Dynamic testing of grandstands and
seating decks, is provided in Appendix 4.

5.2 Aims of testing


Testing may be undertaken with one or more of the following aims in mind.
• To check calculated values of natural frequency used to determine the acceptability
of a grandstand according to the Route 1 method. The programme of testing can also
be designed to reveal whether the most critical mode of vibration has been identified
in the analysis and whether the influence of global modes of vibration needs to be
investigated. (See Section 4.4 on Relevant Natural Frequency).
• To check values of modal properties against those derived using the analytical
model and so determine whether calculations of performance using the Route 2
approach can be considered relevant to the as-built structure. If the discrepancies
between calculation and testing are significant, the results from testing can be used
to inform revisions of the structural model and estimates of dynamic performance.
• To demonstrate to the Client that the assumptions used in calculating natural
frequencies or calculating performance are consistent with the grandstand meeting
the acceptance criteria for the specified Design Event Scenario.
• To monitor the performance of the grandstand under crowd loading during concerts
and other high energy events.

5.3 Circumstances requiring testing


In the absence of records of earlier testing to determine natural frequencies, and unless the
Listed Engineer recommends that testing is not necessary, the following structures should
be tested.
• Grandstands with seating decks that are to be used with pop-concerts and other
events, such as some political, sporting or religious events, where high energy
synchronised rhythmic crowd movement can be expected (Scenario 3 and 4).
• Grandstands where significant complaints have been received concerning motion
experienced in the stand.
• Grandstands where there is to be a change of use to one involving significantly
greater dynamic crowd activity.

8 Dynamic performance requirements for permanent grandstands subject to crowd action


The tests are to be undertaken on the grandstand empty of people but fully fitted out with
seating and services as would be in place during operation.
In planning testing, it will be convenient for the Client, and also good practice, to
include a programme of testing in the acceptance procedures for new and substantially
modified grandstands before handover to the Owners or Managers. The requirements for
testing should be linked to the specification of the stand and the Design Event Scenario
used in design. (See Section 6.5 on Handover procedures for new and structurally modified
structures).

Dynamic performance requirements for permanent grandstands subject to crowd action 9


6  Management responsibilities

6.1 Overall responsibility


The safety of people using viewing facilities such as grandstands is the responsibility of
the Owners and Managers of the facility.

6.2 Design of new stands


With a new grandstand, it is the Client/Management’s responsibility to make sure, from
the outset, that the Design Team includes a Listed Engineer who will have particular
responsibility for advising on dynamic performance. Management’s responsibility then
continues through subsequent discussions with the Design Team in which the use of
the structure, and how it will be operated, is discussed and agreed. Whilst working as a
member of the Design Team, the Listed Engineer also reports directly to the Management
on matters affecting dynamic performance, including any implications that might arise
from changes proposed following ‘value engineering’ to meet budget constraints.
In principle, all grandstands could be required to be designed to provide both safety
and comfort for all possible uses including those likely to produce the most severe dynamic
crowd loading that can be envisaged. In many cases, depending on the use of the stands,
such designs would not be practical or economic. They would have very high initial cost
and would provide levels of performance that might never be needed in the life of the
structure.
The approach recommended here is for Management to be fully involved in
developing operational strategies to be used with the selected Design Event Scenarios
that become the agreed basis for design for dynamic crowd loading. Besides the physical
characteristics of the grandstand, overall dimensions and capacity, each Scenario should
include the descriptions of the crowd and levels of activity to be used by the Design Team
in estimating performance together with any crowd control measures that are required to
be implemented by Management. It is important that the Management understands its role
in setting this agenda for design and the subsequent continuing responsibility to make
sure that the control measures anticipated at the design stage are made effective during
operation of the stand.

6.3 Change of use and assessment for specific events


Management has the responsibility to engage a Listed Engineer and any necessary support,
to form an Assessment Team when changes of use of a grandstand are considered that
could involve the potential for increased dynamic crowd loading. Here the team is led by
the Listed Engineer who reports directly to Management.
Management should accept that the Listed Engineer is engaged to provide an
objective assessment and not necessarily to approve any proposed arrangement. The
Listed Engineer will assess schemes that are proposed and, where this may be helpful,
propose additional measures to mitigate dynamic crowd action and its effects. However,
there could be situations where the Listed Engineer’s advice is that a grandstand should
not be used for certain types of events.
It is important that Management should bear in mind the Design Event Scenario
already used in design or established in a prior assessment and seek the Listed Engineer’s
advice, before proceeding to schedule an event and book a particular Group or Performer.
It will be evident that assessment for specific events should not be overshadowed by
contractual arrangements in which a Music Group or Performer has already been engaged

10 Dynamic performance requirements for permanent grandstands subject to crowd action


to appear. In such circumstances, the Listed Engineer should make sure that the terms
of engagement preclude any liability to meet the costs of a cancellation that might be
considered a consequence of the Listed Engineer’s advice.

6.4 Operational strategies to reduce dynamic response and crowd alarm


The following are examples of management strategies that can be used to reduce the effects
of dynamic crowd loading.
• Netting off. Part of a stand, usually the front rows of a cantilever where dynamic
loading has the most severe consequences, can be made inaccessible to spectators
by ‘netting off’ the relevant rows.
• No standing areas. Areas can be designated where spectators are required to
remain seated. This would avoid the occurrence of most severe forms of dynamic
crowd loading emanating from that area. This strategy requires practical stewarding
issues to be addressed because dynamic response could build up quickly were a
group of spectators suddenly to become active. For Lively Concerts or Pop-concerts
(Design Event Scenarios 3 and 4 in Table 1), the area of designated seating should
have specific ticketing arrangements and the seated people should have an adequate
view of the stage so that any inclination to rise for a better view is eliminated. The
sight lines for designated areas should be assessed on the basis that people in the
remainder of the stand may be standing, bobbing or jumping. People in designated
seating areas should be advised that they are likely to feel motion of the stand.
• Temporary supports. If demonstrated by appropriate calculations or past
performance, temporary supports (or props) can be used to modify the dynamic
behaviour of the grandstand for specific events. Ideally, the potential use of
temporary supports should be considered at the scheme design stage when it will be
easier to arrange for the supports to be associated with an adequate load path.
• Curtailment of music. Arrangements can be made to cut the visual and/or audio
stimulus to crowd behaviour if the structural response approaches an unacceptable
level. This strategy provides an ultimate safeguard but there are difficulties in
implementation. Instrumentation is needed to monitor acceleration of the seating
deck. The build up of response can be very rapid so that there is a need for a fixed
response level for curtailment rather than rely on ad hoc decision making. However,
automatic curtailment could be triggered by a single event or sharp transients that have
little relevance to the overall reaction of the crowd. One approach used in practice
employs a traffic light system on stage with green indicating safe operation, amber
showing caution and red indicating danger which should be followed by cut-off.
• Advice to ticket holders. Experience has shown that it is beneficial to advise the
audience that,
a) during a concert, they may become aware of movement of the stand,
b) some structural movement due to crowd action is expected and considered in
the design, and
c) the stand is designed and operated to be safe and not fail structurally even
under extreme movement.
Experience suggests that advance information of this kind provides reassurance,
minimises the likelihood of complaints and reduces a tendency to panic.

Dynamic performance requirements for permanent grandstands subject to crowd action 11


6.5 Handover of new and structurally modified grandstands
Where testing is included in the specified handover procedures for a new or significantly
structurally modified grandstand, Management, representing the Client, should require the
Listed Engineer, in association with the Practice responsible for the design, to:
• Measure dynamic properties to determine the relevant vertical natural frequencies
and their associated mode shapes and damping values for all relevant areas of seating
in the completed structure. These areas of seating are likely to include long back
spans and cantilevers on long props to the rear of large seating decks as well as front
cantilevers.
• Review the test results and corresponding values calculated using the analytical
model. If necessary, refine the analytical model and compare test and recalculated
results. If not already fully allowed for in the calculation of the dynamic properties,
consider the effects of the actual values of material properties in the as-built structure,
actual connection stiffnesses, foundation stiffness and other boundary conditions used
in the analysis (such as the interface with the roof).
• Report on the results of testing and calculation of dynamic properties.
• For seating decks designed using the Route 1 method, advise Management on the
permissible range of use as given in Figure 1. Alternatively, reassess the design in
the light of the test results using the Route 2 method and advise Management on the
appropriate Design Event Scenario and associated management requirements.
• For seating decks designed using the Route 2 approach, confirm that the measured
dynamic properties are consistent with the performance calculated to meet, or
improve on, the specified Design Event Scenario. Alternatively reassess the calculated
performance in the light of the measured properties and advise Management
accordingly.

The test results and results of calculations should be fully documented and included in an
interpretive report to be included in the grandstand Operations Manual together with detailed
structural drawings of the grandstand.
Management should review the report and recommendations provided with the Listed
Engineer/Design Team and agree any changes necessary to achieve acceptable performance.
These should be documented and included in the Operations Manual.
It should be noted that it is the responsibility of Management, acting for the Client,
to make sure that consequences of the measured dynamic performance and any changes
in the Design Event Scenario and associated management requirements are communicated
to the Local Authority safety advisory group and all operational personnel including those
planning events, police and other emergency services, safety managers, ground staff and
stewards as necessary.

6.6 Operations Manual


Management should not accept a new or structurally modified stand without being provided
with the Operations Manual that has been prepared for the stand by the principal contractor
with assistance from the designers where required.
The following should be included in the Operations Manual:
• A summary of the dynamic response characteristics including lowest relevant vertical
frequencies, associated mode shapes and damping for each significant area of seating
in the structure incorporating the stand.

12 Dynamic performance requirements for permanent grandstands subject to crowd action


• A description of the Design Event Scenarios considered and the anticipated response
of the structure to these together with any crowd management measures required to
be implemented by Management.
• A record of the Engineering Practice and Listed Engineer responsible for the
static analysis of the structure and of the dynamic analysis of the structure and
the Test Agency responsible for the measurement of dynamic performance. The
record should include similar records for any separate sub-analyses of the dynamic
performance of elements within the structure carried out by others such as for
precast concrete seating units.

The Operations Manual should be up-dated following changes in operational procedures


and/or structural modifications.

6.7 Operation
The assessment of the dynamic performance of a grandstand should be seen as part of the
broader risk assessment dealing with the event, the venue and the crowd and the implications
for safe management.
In relation to the structure itself, it is accepted that, with appropriate management
and controls, many existing stands can be operated safely even though the relevant natural
frequency does not satisfy the Route 1 requirements. However, if a stand is to be used in
these circumstances, it is important that control measures adopted are based on the following
principles.
• An existing agreed and recorded Design Event Scenario which should include details
of any measures adopted to reduce dynamic response. (See Section 6.4).
• Sufficient knowledge of the relevant dynamic properties of the structure and the
behaviour of the stand under earlier, and possibly less severe, conditions of crowd
loading.
• Use of a Listed Engineer for detailed assessment and direction on measures needed to
implement Route 2 Recommendations.

In addition, Management should be aware that both design and assessment for dynamic
crowd loading involves uncertainties, particularly in the make up of a crowd at any particular
event and the level of excitation that the crowd provides to the structure. Because of this, the
crowd management requirements included in the Design Event Scenario should be seen as
good guidance based on the best knowledge available but subject to review following each
event.
It is within the responsibilities of Management to build a knowledge base concerning
the performance of its grandstands so that management controls can be revised if this is
necessary to maintain adequate safety levels. Fine tuning of crowd control measures can be
expected following the first use of a stand for different types of events.
The knowledge base should typically contain:
• A detailed description of each event.
• CCTV records of crowd activity preferably synchronised with an audio record.
• Audio tapes from concerts; particularly pop-concerts and high energy events.

Dynamic performance requirements for permanent grandstands subject to crowd action 13


• Stewards’ reports noting any signs of structural motion and extremes of crowd
behaviour including signs of panic. – a standard report form is available and can be
downloaded from www.istructe.org/technical/db/277.asp
• Records of complaints, relating to perception of structural motion, from members of
the public attending events. (See Section 8.3 on individuals’ tolerance of motion).

These operational records should be available for reference by the Listed Engineer when
employed to advise on modifications to existing procedures or structural modifications.

Management to monitor f0 > 6


use of incidental music Suitable for all
and maintain records of types of events
audience feedback Scenarios
including complaints 1, 2, 3 and 4

For seating areas with


f0 < 3, resonance f0 > 3.5
could occur with the Minimum for new construction
first harmonic of the Suitable for viewing sport and other events
crowd loading with with predominantly seated audiences.
consequent large Scenarios 1 and 2
structural movements

Management
is advised to Existing stands with 3 < f0 < 3.5 may be deemed
monitor the Only satisfactory for sports viewing (i.e. Scenario 1) on
crowd’s Scenario the basis of past experience and use for less lively
reactions at 1 sections of the crowd
all events

2 3 4 5 6 7

Relevant vertical natural frequency of seating deck, f0 Hz

Figure 1  Route 1 requirements for different categories of use

14 Dynamic performance requirements for permanent grandstands subject to crowd action


7  Route 1: Compliance with natural frequency requirements

7.1 Vertical excitation


Grandstand seating decks should meet the requirements of relevant natural frequency
for different categories of use as shown in Figure 1.

7.2 Side-to-side horizontal excitation


In contrast to temporary or demountable grandstands, there has been little cause for
concern over the behaviour of existing permanent grandstands subject to horizontal
excitation due to crowds with individuals swaying from side to side or indulging in
a Mexican wave. No problems associated with crowd comfort should be experienced
for grandstands with natural frequencies for horizontal excitation greater than 1.5 Hz.
However, the connection of the seating deck to the principal members and also the
primary structure should be designed to resist the horizontal loads due to side-to-side
motion of a crowd. Accordingly, it is recommended that the stand should be designed
to withstand the additional horizontal loads in Table 2 as part of the design for static
loading.

7.3 Nodding modes due to front-to-back excitation


Because of the rake of a cantilever seating deck, there will be a horizontal component of
displacement as well as a vertical component due to crowd action. This is accentuated
if the structure supporting the cantilever is itself flexible. The result is a nodding mode
encompassing both the cantilever and its support. This behaviour is distinct from side to
side motion due to sway of crowd and is primarily a result of vertical excitation with the
response being magnified due to lack of stiffness in the structure supporting the seating
deck. (See Appendix 1, Section A1.5.2 on acceleration limits).
As in the treatment of side-side horizontal excitation, it is recommended that the
stand should be designed to withstand the additional horizontal loads in Table 2 as part
of the design for static loading. The occurrence of nodding modes and their significance
should also be investigated as part of a Route 2 analysis.

Dynamic performance requirements for permanent grandstands subject to crowd action 15


Table 2 Design for horizontal strength and stability
To be used with both Route 1 and Route 2 methods of design and assessment
In addition to the operational wind loading, grandstands should be capable of
withstanding the following lateral loading due to crowd action. The loads should be
incorporated in the static design for ultimate load of the structure in combination with
other design loads.
Type of use Additional static horizontal load as a
percentage of the specified static live
loading on the seating deck
Side to side Front to back
All grandstands except those used for ±5% ±5%
pop-concerts or similar lively activity
Grandstands to be used for pop-concerts ±7½% ±7½%
or other lively events
Notes
i) The loads are specified as a percentage of the specified live loading on the
seating deck. Note only loading on the seating deck needs to be considered.
ii) The horizontal load should be applied in the plane of the seating deck in the way
the people are situated according to the available seating.
iii) The horizontal loads due to crowd action are additional to loadings from other
causes and so should be applied in combination with operational wind loads.
iv) The partial load factors to be used in each load case should be those specified
for live loads in the appropriate Code of Practice for the structural material
involved. A partial factor of 1.5 should be used with the given horizontal loads in
load combinations with factored values of dead and imposed loads.

16 Dynamic performance requirements for permanent grandstands subject to crowd action


8  Route 2: Design for managed events

8.1 Outline
The Route 1 design method is intended to provide a simple check for safety and serviceability.
The approach is based solely on the physical characteristics of the grandstand and so
does not require analysis of the performance of the grandstand under dynamic crowd
loading or consideration of particular measures of crowd control that might be adopted by
Management.
The Route 2 method requires the grandstand to be analysed to estimate its performance
under dynamic loadings specified for different classes of activity and size of crowd on
the seating deck as described in the relevant Design Event Scenario incorporating the
management controls. The recommended method of analysis requires consideration to
be given to human structure interaction due to the grandstand acting in combination with
the crowd treated as load generating structural elements. (See Appendix 1 giving the
background to human structure interaction).
The performance of the seating deck is described by the displacements, accelerations
and stress resultants calculated using the specified crowd loading. The grandstand’s
performance should be assessed on the basis of the given requirements for serviceability
and ultimate load capacity.
For all except the most flexible structures, integrity checks for ultimate load
capacity are likely to be relevant only to connections between structural elements. Also,
displacements can be expected to be within acceptable limits provided acceleration limits
related to tolerance of motion are not exceeded.

8.2 Idealised description of crowd activity


People attending an event do not sit or stand without moving. All individuals in a crowd
are likely to move, and if music is played, the movements tend to become synchronised
to the beat. The movements may be an involuntary reaction to the external stimulus and
the behaviour of neighbouring people in the crowd. Even at this level, there will be some
dynamic response that will be felt by individuals if the grandstand is unduly flexible.
However, with events such as pop-concerts, there is an expectation of excitement. The
crowd expects to be involved; participation is deliberate and individual motion may
become extreme with foot-stamping, bobbing (sometimes termed bouncing) and possibly
some jumping on the stand, all in time to a beat.
For the purpose of design, sections of a crowd are idealised as being either
predominantly,
• Standing with dynamic properties given in Appendix A2 and generating loading
due to body motion over a wide frequency range including foot stamping, bobbing
and a proportion of jumping (i.e. active), or
• Seated with the dynamic properties given in Appendix A2 and regarded as inactive
(i.e. passive).

If areas of designated seating are used to reduce the overall level of excitation, these must
be agreed with Management and controls exercised as outlined in Section 6.4.

Dynamic performance requirements for permanent grandstands subject to crowd action 17


8.3 Serviceability: Tolerance of motion
For the purpose of design, the maximum root mean square (RMS) acceleration of the
seating deck is used as a measure of what is felt by people in the stand.
Maximum RMS acceleration limits are given in Table 1 for the different Design
Event Scenarios and are to be used with the loadings recommended in Appendix 2 for
Route 2 calculations of performance related to design or assessment.
The design acceleration limit for Scenario 4 is regarded as the maximum that can be
tolerated without the prospect of panic by individuals in a crowd.
For Scenarios 2 and 3, the level of comfort to be provided is essentially a matter for
the grandstand Management. The recommended crowd loading is weighted according to
the expected occurrence of song frequencies over a number of events and the acceleration
limits reflect the situation that seated people will be more sensitive to motion than people
who are standing or moving in time to a rhythmic stimulus.
The tolerance of motion varies between individuals and will not be the same for
different style events. Design according to a given Event Scenario will not necessarily
guarantee that all individuals will react with the same degree of satisfaction to the level of
comfort provided or that there will be a total absence of complaints.
More stringent requirements may be appropriate for some sections of a grandstand
or seating deck such as catering areas and hospitality suites where expectations for comfort
and operational convenience will be at a premium.

8.4 Serviceability: Displacement limits


The maximum dynamic component of displacement due to crowd loading should not
exceed 7mm RMS.

8.5 Ultimate load capacity


No dynamic loading check is required of ultimate load capacity for grandstands designed
and managed for Scenarios 1 or 2 since the ultimate load capacity required by current UK
standards for dead and imposed loading for human occupancy, combined with the additional
horizontal imposed loads specified in Table 2, will always be sufficient. The same is true
for grandstands designed and managed for Scenario 3 or 4 other than those where motion
of the deck is significantly affected by a global mode involving the supporting structure
for which a separate check may be advisable.

8.6 Fatigue
For stadia with frequent use for Scenario 3 and 4 events, the possibility of fatigue damage
may need to be considered. The precise use of a stand over time can seldom be anticipated
but an initial approximate indication of the potential for fatigue can be based on 20 minutes
exposure per concert to Scenario 3 loading. This estimate should be revised as the pattern
of use develops over time or in response to observed behaviour.

18 Dynamic performance requirements for permanent grandstands subject to crowd action


9  Analysis of dynamic performance

9.1 Impulse loads


No recommendations are made for single sharp increases in load due to sudden
movements of members of a crowd as might occur when a goal is scored at a football
match. Such events may cause motion that can be felt by members of a crowd but are
normally of short duration and so of little significance for crowd comfort. Of much
more importance is periodic loading induced by crowd activity coordinated to coincide
approximately with a given frequency, possibly promoted by an external stimulus such
as a musical beat. In this context, the loading may be more or less severe depending
on the freedom of movement of people within the crowd and whether they are seated,
standing, bobbing or jumping.

9.2 Horizontal loads due to periodic excitation


Until more complete information becomes available, horizontal loads due to crowd action
should be treated in the same way as in the Route 1 method so that Grandstands should
be designed to resist the equivalent static loads given in Table 2. (See also Sections 7.2
and 7.3).
It is hoped that, in time, sufficient information will become available for horizontal
actions to be dealt with using a human structure interaction based dynamic analysis as is
recommended for vertical loading.

9.3 Analysis for vertical periodic excitation


Linear elastic dynamic analysis with linear damping should be used. In practice, except for
the simplest structures and crowd configurations, this will involve using modal analysis
and a finite element representation of the structural system. Additionally, the effects of
human structure interaction need to be included in the analytical process whatever method
is adopted. (See Appendices 1 and 2).
Except where more precise information is available from testing the as-built structure,
damping should be taken to be 2% critical for each mode of vibration considered.

9.4 Human structure interaction


Recent research has shown that people involved in any of the three activities of sitting,
bobbing or jumping interact with moderately flexible structures such as grandstand seating
decks so that the contact forces actually experienced at resonance are significantly different
from those measured in tests on stiff or rigid force plates. The existence of these effects does
not depend on the motion of the structure being extravagantly large but only on the relative
stiffness and damping properties of the crowd and the supporting structure. If the effects
due to human structure interaction are ignored in calculations, the response of the structure
will be incorrectly represented in the analysis. For most practically designed grandstands,
accelerations calculated ignoring human structure interaction will be significantly higher
than those calculated taking into account the interaction effects.
The recommended analytical method for treating human structure interaction is
given in the Appendices. This approach has been developed, using the most recent research
and experimental data available, with the aim of reproducing the patterns of behaviour
observed in actual structures subject to dynamic crowd loading. The method cannot deal
with all the variations in human behaviour and physical characteristics that affect the
way in which individual people and crowds interact with a structure. Because of this,
the results of analysis should be regarded as indicative of actual behaviour rather than a

Dynamic performance requirements for permanent grandstands subject to crowd action 19


precise prediction of performance for a particular event. More positively, the recommended
methods provide a consistent approach to design and assessment that takes account of the
major factors that determine the dynamic behaviour of a stand.
The Appendices also provide the necessary input concerning loads appropriate to
the different Design Event Scenarios together with notes on testing, calculation of natural
frequencies and analysis.

20 Dynamic performance requirements for permanent grandstands subject to crowd action


10  Use of the Recommendations

The Recommendations are addressed to both Management and the Design Team. The
Recommendations provide a guide to operational management and a tool for the analysis,
design and assessment of grandstand seating decks for dynamic crowd loading. The
Recommendations have been prepared on the basis that these are related functions with
Management’s role in operating a grandstand having importance similar to that of the
Design Team that provides the details for the stand’s construction.
The Recommendations represent the most considered view now available for the
treatment of dynamic crowd loading on seating decks of permanent grandstands. This is
now an area for continuing research and it is likely that the detailed recommendations for
analysis will be refined over time. It would be helpful if users of the Recommendations,
and researchers, who have relevant material to contribute, would provide the Institution of
Structural Engineers with comment so that the Recommendations can, from time to time,
be refreshed in the light of experience and new knowledge.

Dynamic performance requirements for permanent grandstands subject to crowd action 21


Appendix 1  Background to human structure interaction

A1.1 Introduction
Explicit treatment of human structure interaction has not been previously considered
in design guidance for active crowd loading. In the past, it has been assumed that the
load induced by an individual, or a crowd, is an externally applied load unaffected
by the motion of the structure. This assumption led to loads obtained in tests using
people bobbing or jumping on relatively stiff structural elements being presented in
recommendations as being appropriate for all situations. For grandstands with dense
crowd loading and natural frequencies typically less than 7Hz, this approach gives
insufficient consideration to the nature of the loading – due to an individual or a crowd
– or to the effects of the mechanical interaction between individuals and the structure.
These aspects have been recently addressed (Dougill et al., 2006) using a simple structural
model for the active crowd that interacts with the structure during motion. Laboratory
based studies have demonstrated the significance of this interaction for a range of support
natural frequencies, loading and excitation relevant to grandstands (Yao et al., 2004 and
2006). Also, use of the theoretical model (Pavic and Reynolds, 2008), and independently
derived loading data (Parkhouse and Ewins, 2006), has allowed the performance of
actual grandstands to be calculated and compared satisfactorily with observed data from
stands in service (Pavic and Reynolds, 2008). Both the laboratory tests and full-scale
studies have shown that, for practically designed grandstands and dense crowd loading,
it is necessary to take account of crowd-structure interaction if the structural response
near resonance is not to be significantly overestimated.

A1.2 Basic principles


A1.2.1 Modelling human structure interaction
In calculating the relevant dynamic properties of the empty seating deck, a linear elastic
representation of the structure will have already been developed, for example using
a finite element model. In order to represent the effects of crowd loading, this basic
structural model is supplemented by additional elements representing groups of people.
These crowd elements, or body units, are spring-mass-damper systems, as shown in
Figure A.1.1, each energised by an actuator represented by the forces P(t) that cause
dilation of the crowd or body unit by means of a pair of equal internal forces applied
in opposite directions. The properties of the crowd units depend on whether people are
predominantly standing or sitting whilst the forces P(t) relate to the type and intensity
of activity in the crowd. In the context of analysis, the body units, with their associated
forcing functions, replace the forces that have commonly been prescribed in design
guidance to represent crowd loading. Use of the body units provides a model of real-life
crowd loading in which motion of the combined structure/crowd system is caused solely
by forces generated within the system itself.
The motion of each crowd body unit is determined by the relative displacement of
the body mass with respect to the point at which the unit is in contact with the structure. As
a result, each crowd body unit introduces an additional unknown degree of freedom into
the structural system associated with the mass of the crowd body unit.
The forces, P(t), are taken to be periodic. For practical periodic crowd loading, the
function can be represented by the first three harmonic components. It follows that the
combined structure/crowd system can be analysed using conventional methods of linear
dynamic analysis to determine the separate responses due to each load harmonic. The
separate responses can then be combined to obtain the total response to crowd loading.

22 Dynamic performance requirements for permanent grandstands subject to crowd action


Body unit with mass,
spring stiffness and Crowd body node with
damping associated nodal displacement

Internal force pair


driving the system P(t)

Basic structure node with


associated nodal displacement

Contact force with the structure

Figure A1.1  Typical body unit for incorporation into the basic structural
model of the supporting structure empty of people

Dougill, Wright, Parkhouse and Harrison (2006) provide the governing equations for a
system comprising a single degree of freedom structure energised by a single body unit.
Formal solutions are given for harmonic loading – P(t) being either a sine or cosine
function of time – together with examples of the resulting behaviour of the combined
system. In practice this would correspond to the situation when the unoccupied structure
has a single dominant mode and so can be taken as a single degree of freedom system for
the purpose of dynamic analysis. If the crowd is relatively homogeneous, it can be defined
by a single crowd body unit resulting in a combined two degree of freedom (2DOF) system,
as described in A1.5.1.

A1.2.2 Active and passive behaviour


Interaction between a moving structure and a crowd occurs for both active people, who
cause the structure to move by their own efforts through the driving forces P(t), and passive
people, who do nothing themselves to cause motion of the structure and for whom P(t)=0.
Besides the difference in loading, the dynamic properties of the body units also differ
between those representing active and passive people.
The recommended standard Design Event Scenarios (given in Table 1) are based on
limiting conditions for various types of events in which everyone is considered to be active
so that there is no passive contribution.
If designated seating areas are specified in the design or as an agreed operational
strategy prior to assessment (as described in Section 6.4 No standing areas) the relevant
part of the structural model would need to be combined with the appropriate passive
elements in addition to using active elements for those parts of the crowd considered to
cause the structure to move.

A1.3 Application of the theory


A1.3.1 Direct application of the theory
In principle, the most straightforward and accurate way to use the human structure
interaction theory is to add crowd body units to the finite element model of the empty
structure and to analyse the combined system. No assumptions need be made, beyond
those involved in determining the loading and body unit properties and ensuring that the
internal damping of the empty structure is suitably modelled, so that the results obtained
are analytically correct for the chosen distribution of body units.

Dynamic performance requirements for permanent grandstands subject to crowd action 23


Use of the direct approach would allow a comprehensive analysis of the effect of
a particular crowd on a structure. For example, body units, with different properties and
loadings, could be used to represent different groups of people on the seating deck, or
decks, with different levels of activity including passive behaviour. This amount of local
detail is normally unnecessary in design but enough body units need to be used so that the
motion of the crowd is adequately represented.

A1.3.2 Approximate analysis using an assumed mode shape for the


crowd’s motion
An approximate solution can be obtained using an assumed ‘mode shape’ for the
displacement of the crowd (and its associated body units) for each mode of the empty
structure that is being considered for the seating deck or decks. The modal masses,
stiffnesses and internal drivers, P(t), can then be found using these assumed mode shapes.
The structural response then follows through the usual processes of modal analysis using a
two degree of freedom model for each relevant mode of the empty structure.
The most convenient approach is to assume that the body units adopt mode shapes
that are identical to those of the empty structure in their vicinity. This assumption appears
to be borne out by observations on cantilever decks and can be shown to be exact for
simple structures with constant section properties, fully occupied by a uniform crowd with
each part of the crowd having the same properties. This has led to successful correlation
between calculated and observed motion for plate-like cantilever grandstands (Pavic and
Reynolds, 2008).

A1.4 Body Unit properties and loadings


Recommendations for the body unit properties and loadings appropriate to the different
Design Event Scenarios are given in Appendix 2.

A1.5 Analysis and results


A1.5.1 Modal analysis
The most usual form of analysis will be using modal decomposition. In essence this allows
the separate contributions of each mode to the overall response to be analysed separately
and then combined using the principle of superposition. The process involves identifying
the modes of the unoccupied structure that contribute to the motion of the deck, together
with their associated mode shapes, and then calculating the relevant modal properties for
a two degree of freedom (2DOF) crowd-structure system that can be derived for each
mode using the mode shape of the unoccupied structure and the distribution of mass and
dynamic loading within the crowd-structure system. (See Appendix 3). For a structure, fully
occupied with an active uniform crowd, this leads to a pair of equations corresponding to
each mode that describes the motion of the structure itself. These can then be solved to
obtain the body unit’s and structure’s response as they relate to that particular mode. A
formal solution of these equations for periodic loading is available (Dougill et al., 2006)
that can be used to benchmark numerical methods used in analysis

A1.5.2 Root mean square (RMS) accelerations and acceleration limits


Both BS 6841(1987) and ISO2631-1-(1997) use Root Mean Square (RMS) acceleration as
the indicator for assessing tolerance of motion. For any function x(t) the RMS value over
the interval of time T is given by,
t+T 1/2

RMS ^ x, t, T h = > T # x dt H 
1
(A.1.1)
2

24 Dynamic performance requirements for permanent grandstands subject to crowd action


When monitoring actual movements, the underlying periodic motion due to crowd activity
is accompanied by irregular events. The RMS value therefore varies with time and depends
somewhat on the choice of the interval T. A value of 10 seconds is frequently used leading
to ‘10 second rolling RMS’ values over the period of observation.
Calculations used in design based on periodic motion using the harmonic loading
described in Appendix 2 lead to accelerations that are smooth functions of time. T is taken
to be the excitation period for the harmonic considered leading to RMS values that are
independent of time. For a single harmonic with frequency f and amplitude a, such that
x = a cos(2πft), the RMS value of the acceleration x is then a/√2.
The BS and ISO Standards additionally weight measured RMS acceleration according
to excitation frequency. This has not been considered appropriate for the values calculated
for the range of frequencies encountered with crowd motion. Accordingly, the acceleration
limits in Table 1 of the Recommendations use RMS accelerations calculated for periodic
motion and without frequency weighting.
For most situations it will be sufficient to consider only the vertical component
of acceleration in meeting the limits for acceleration given in Table 1. However, with
‘front-to back’ nodding modes (See Section 7.3) there is the prospect that the horizontal
component of acceleration will be significant even with only vertical excitation. The check
for acceptable motion should then be based on the limits in Table 1 and the vector sum of
the calculated vertical and horizontal RMS accelerations.

A1.5.3 Analysis with a dominant mode.


In general the analysis should cover all modes that are considered to be capable of
providing significant motion of the seating deck. (See Section 4.2). However, in some
situations there will be a dominant mode so that the response of the crowd/structure
system can be determined from a single pair of equations corresponding to the dominant
mode. In general, three loading harmonics will be considered corresponding to excitation
at the activity frequency, f, and the resulting higher harmonics with frequencies 2f and 3f.
The acceptance criteria in the Recommendations are expressed in terms of the maximum
permitted Root Mean Square, RMS, accelerations of the structure (see Section 2, Table
1) so that the combined effect from the three harmonics must be determined. This can be
done for any given activity frequency from,

RT = R 21 + R 22 + R 23  (A1.2)

where R1, R2 and R3 are the RMS values for the response (acceleration or displacement
as required) due to the 1st , 2nd and third harmonics of the activity frequency and RT is the
RMS value of the total response at that frequency.
In doing the analysis, it should be recognised that the maximum response will not
necessarily occur at the natural frequency of the unoccupied structure or even at a natural
frequency of the combined crowd/structure system due to frequency dependence of the
specified body loading. Accordingly, the analysis will need to include a frequency scan,
with results being obtained over a range of closely spaced frequencies, in order to identify
the maximum response.

A1.5.4 Multi-mode analysis


In general, more than one mode will need to be considered. Again a frequency scan should
be used to determine the RMS responses for a range of frequencies for each loading
harmonic and each mode. The results from the separate modes can be combined in a
similar manner to that indicated in equation A1.1 with R 1 , for instance, being the sum
2

Dynamic performance requirements for permanent grandstands subject to crowd action 25


of squares of the RMS contributions to the first harmonic response from all the modes
considered.

A1.6 References
Yao, S., Wright, J.R., Pavic, A. and Reynolds, P. ‘Experimental study of human-induced
dynamic forces due to bouncing on a perceptibly moving structure’, Canadian J. Civil
Engineering, 31(6), 2004, pp1109-1118.
Yao, S., Wright, J. R., Pavic, A. and Reynolds, P. ‘Experimental study of human-induced
dynamic forces due to jumping on a perceptibly moving structure’, J. Sound & Vibration,
296, 2006, pp150-165.
Parkhouse, J.G. and Ewins, D.J. ‘Crowd induced rhythmic loading’, Proc. ICE, Structures
and Buildings, 159(SB5), Oct 2006, pp247-259.
Dougill, J.W., Wright, J.R., Parkhouse, J.G. and Harrison, R.E. ‘Human structure interaction
during rhythmic bobbing’, The Structural Engineer, 84(22), 21 Nov 2006, pp32–39.
Pavic, A. and Reynolds, P. ‘Experimental verification of novel 3DOF model of grandstand
crowd-structure dynamic interaction’. 26th International modal analysis conference:
IMAC-XXVI, Orlando, Florida, 4-7 Feb 2008, paper 257.

26 Dynamic performance requirements for permanent grandstands subject to crowd action


Appendix 2  Body unit properties and recommended loading

A2.1 Body unit and structure


Figure A2.1 shows a typical body unit in place on the supporting structure. The forces
produced within the body unit which induce motion are shown as P(t). The resulting
contact force on the support is F(t). The structure is not shown in detail. If the structure
were rigid, the contact force would equate to the force corresponding to conventionally
determined Dynamic Load Factors as used in analysis ignoring the effects of human
structure interaction.
In this Appendix, recommendations are given on values of the body unit properties
and internal forces for direct use in design or assessment using this human structure
interaction model.

A2.2 Crowd body elements


For bobbing that is comfortable and easily maintained, Parkhouse and Ewins’ (2006)
results led to the conclusion that the body properties could be taken as independent of
group size. (Dougill et al., 2006).
For the purpose of analysis, only two types of crowd body element are proposed.
These comprise the mass/spring damper system shown in Figure A2.1 with the properties
given in Table A2.1

Body unit with mass,


spring stiffness and Crowd body node with
damping associated nodal displacement

Internal force pair


driving the system P(t)

Common node with body unit


Contact force with
the supporting structure
F(t) and structure with associated
nodal displacement

Figure A2.1  General arrangement of body unit and supporting structure

Dynamic performance requirements for permanent grandstands subject to crowd action 27


Table A2.1. Recommended properties of crowd body elements
Crowd Event scenario Natural Damping
frequency percent
Hz critical
Designated and controlled See Recommendations 5 40
‘no-standing’ areas Section 6.4
Predominantly Scenario 2 5 40
seated
Active and mostly standing Event Scenarios 3 and 4 2.3 25

The body spring stiffness, k, is found from the body mass, m, and the natural
frequency, n, as given in Table A2.1, from,

k = 4r2 n 2 m  (A2.1)

The linking member with the basic structure marked in Figure A2.1 as containing the
common node with the structure can be regarded as rigid but with no mass. The concentrated
mass, m, in the element should be calculated from the number of people in the area of
seating deck that affects the node of the structural model to which the element is attached,
using an average person mass of 80kg.

A2.3 Representation of periodic loading


The internal periodic force pair P(t) in figure A2.1 is described as the sum of by three
harmonic components as follows,

 (A2.2)

where
t is the crowd effectiveness factor that reflects design criteria driven
primarily by serviceability with commonly occurring events for Scenarios
2 and 3, or mitigation of the potential for panic under extreme motion
considered in Scenario 4.

m is the mass of the crowd associated with the particular body element
considered. This is to be taken as 80 kg times the number of people.

g is acceleration due to gravity 9.81 m/s2.

Gi is the ith ‘generated load factor’ GLF defining the load generated by
activity of the crowd.

f is the fundamental frequency of the crowd activity in Hz. This


corresponds to the musical beat (or frequency of an alternative prompt) in
beats per second.

t is the time in seconds.

ii is the phase difference of the ith harmonic. These phase differences can be
set to zero in calculations if only RMS values of force, displacement or
acceleration are required.

28 Dynamic performance requirements for permanent grandstands subject to crowd action


The Generated Load Factors, GLFs, replace the Dynamic Load Factors used in analyses
that ignore the effects of human structure interaction. The differences in response are small
when the crowd is sparse (that is for low crowd mass to structure mass ratios) and for
structures with natural frequencies that are high compared with those of the crowd body
elements. In other situations, human structure interaction needs to be considered.
The principal assumption in developing the GLFs from available data is that the internal
forces P(f,t) used by groups of people to move on a flexible platform will be the same as
would be involved when undertaking the same activity on a rigid base. However, crowds
engage in different types of activity and most research has been concerned with jumping
that can be related more to aerobics and vigorous dancing than to crowd behaviour in stadia
where people react to music, over a wide frequency range, with handclapping, stamping,
bobbing and occasional jumping. Accordingly, for the purpose of design, the loading for
Scenarios 3 and 4 is idealised and taken to be equivalent to a multiple of loading due to
bobbing by groups of 50 people or more modified by an effectiveness factor. The factor takes
some account of activity that might occur over a wider frequency range than is covered by
the testing and, for the commonly encountered events of Scenarios 2 and 3, also provides a
weighting based on the frequency of occurrence of songs with different tempi.
The basic loads, Gi, (i = 1,2,3) before modification by an effectiveness factor,
t, are based on results for bobbing from extensive recent testing by Parkhouse and
Ewins (2006) that also provide data to enable the body unit properties to be determined
(Dougill et al., 2006).

A2.4 Internal ‘drivers’- Gi – producing dynamic crowd loading


The recommended values for the Generated Load Factors are given in Table A2.2. These are
derived from the synthesised results for a group of 50 people bobbing on a rigid platform.
The values can be taken as constant for larger groups of people. For smaller groups, there
is a significant increase in Gi with group size, together with increasing variations from
the mean. In modelling a crowd, it is recommended that the crowd is not subdivided into
groups of less than 50 people.

Table A2.2 Recommended values of the Generated Load Factors, Gi, for use in
calculations of performance for design or assessment.
Harmonic number Effectiveness
Scenario Typical activity represented
i=1 i=2 i=3 factor

Not required. Not required.


Route 1 only, but with discretion
1 ------- ------- -------
available by Listed Engineer
(Recommendations Table 1)
Predominantly seated with Eqn. A2.3
occasional coordinated
2 0.12 0.015 Zero
rhythmic movement from
standing people
All crowd considered active. Eqn. A2.3
Moderate bobbing at three
3 0.188 0.047 0.013
quarters Parkhouse and Ewins’
50 person level
The whole crowd active. Eqn. A2.2
Loading taken to be twice that
4 0.375 0.095 0.026
for the commonly occurring
events of Scenario 3

Dynamic performance requirements for permanent grandstands subject to crowd action 29


A2.5 The crowd effectiveness factor ‘t’.
A2.5.1 Scenario 4
Scenario 4 is concerned with high energy events. The principal concern is safety against the
prospect of crowd disturbance, or panic, as a consequence of the motion of the grandstand.
Experience suggests that the probability of such a situation occurring is small. However,
the consequences of panic in a crowd confined in a grandstand with fitted seats could be
dramatic and possibly life threatening. For such an extreme situation it is appropriate to
consider an unrestricted frequency range of possible excitation but with some allowance
for reduced effectiveness of the loading at low and high excitation frequencies. This is
provided by an effectiveness factor shown in Figure A.2.2 and given by the equation,
t (f) = sech ^ f - 2 h  (A2.3)

where f is the fundamental frequency of the crowd’s activity in Hz and the factor is
used in equation A2.2 with the recommended values of Gi in Table A2.2 and the crowd
body unit properties of Table A2.1.

1.2

1.0

0.8

t(f)
0.6

0.4

0.2

0
0 1 2 3 4 5 6

Activity frequency, f, Hz

Figure A2.2  Effectiveness factor t(f) for use with Scenario 4 in


consideration of safety

A2.5.2. Scenarios 2 and 3


Scenarios 2 and 3 are relevant to less energetic events than those of Scenario 4 so that
the occurrence of panic due to crowd motion can be discounted. It follows that criteria
for structural performance can be set in terms of what is needed to meet a crowd’s
expectations of comfort over a period of time and for many events. This element of
repeated exposure to crowd action and a cumulative experience of comfort relative to
motion of the structure is in direct contrast to conditions for the single extreme event
in which safety is the prime consideration. In order to take account of the continuing
exposure to a variety of songs of different beat frequencies, the effectiveness factor

30 Dynamic performance requirements for permanent grandstands subject to crowd action


for comfort is weighted, using data from Littler (2003), according to the probability
of occurrence of songs in the overall pop repertoire. The probability distribution is
approximately Normal with mean 1.8 and standard deviation 0.5, so leading to the
effectiveness

e
 (A2.4)

The resulting effectiveness factor is shown in Figure A2.3. This covers the beat frequency
range of commonly occurring songs with a maximum near to that for the most frequently
occurring songs, Littler, (2003).

1.2

1.0

0.8

t(f)
0.6

0.4

0.2

0
0 1 2 3 4 5 6

Activity frequency, f, Hz

Figure A2.3  Effectiveness factor t(f) for use with Scenarios 2 and 3 in
consideration of comfort
A2.6 Monitoring and back analysis
Event monitoring and subsequent back analysis of performance is most likely to be
undertaken under conditions similar to Scenario 3. However, in examining a specific
event, a view should be taken of the proportion of the crowd that is actively involved
with the remainder being considered as passive. For most events, the loading from active
people will correspond to that for bobbing. Appropriate values for the internal drivers Gi
are given in Table A2.3. These should be used with the crowd body properties for active
and passive crowd body elements in Table A2.1 and the single event effectiveness factor
of equation A2.3. An example of monitoring and back analysis is provided by Pavic and
Reynolds (2008).

Dynamic performance requirements for permanent grandstands subject to crowd action 31


Table A2.3 Suggested values of the Generated Load Factors, Gi, for use in back
analysis of specific events.

Harmonic Number
Typical activity in identified Effectiveness
Crowd
sections of the crowd Factor
i=1 i=2 i=3

Active 0.25 0.063 0.018 Active crowd, mainly bobbing. Eqn. A2.3

Passive All zero Inactive, standing or seated

A2.7. References
Dougill, J.W., Wright, J.R., Parkhouse, J.G. and Harrison, R.E. ‘Human structure interaction
during rhythmic bobbing’, The Structural Engineer, 84(22), 22 Nov 2006, pp32–39.
Littler, J.D. ‘Frequencies of synchronised human loading from jumping and stamping’,
The Structural Engineer, 81(22), 18 Nov 2003, pp27–35.
Parkhouse, J.G. and Ewins, D.J. ‘Crowd induced rhythmic loading’, Proc. ICE, Structures
and Buildings, 159(SB5), Oct. 2006, pp247-259.
Pavic, A and Reynolds, P. ‘Experimental verification of novel 3DOF model of grandstand
crowd-structure dynamic interaction’, 26th international modal analysis conference:
IMAC-XXVI, Orlando, Florida, 4-7 Feb 2008, paper 257.

32 Dynamic performance requirements for permanent grandstands subject to crowd action


Appendix 3: Calculation of modal properties

A3.1 Introduction
The Recommendations provide alternative routes to design or assessment of a
grandstand.
Route 1 uses natural frequency as an index of quality for a stand and requires
knowledge of the natural frequencies of the stand and identification of the lowest value
corresponding to a mode that can be excited by and felt by people on the seating deck.
The approach is simple but rendered useless if natural frequencies are not determined
to sufficient accuracy. As outlined in the Recommendations, (Section 4), calculation of
natural frequencies may appear to be a straightforward task but, in reality, is complicated
by uncertainties in setting up the analytical model with almost inevitable differences
between assumptions made for purposes of calculation and the actual behaviour of the
as-built structure.
The Route 2 approach provides more flexibility for the designer by using calculation
to obtain an estimate of performance of a grandstand under prescribed loading appropriate
to a given idealised Design Event Scenario. A full dynamic model of the structure is
needed for such calculations and the opportunities for error and miss-match between the
analytical model and real structural behaviour are certainly not less than in calculations of
natural frequencies for Route 1. Clearly, if the calculated dynamic properties are seriously
in error, the resulting estimate of performance will have little relation to the behaviour of
the as-built structure.
To deal with this potentially difficult situation, engineers need to be aware of
the assumptions or simplifications made in analysis and how these affect the result of
dynamic analysis. They should also recognise that even the most careful attention to detail
in analysis cannot guarantee that the analytical model will match the behaviour of the
physical as-built structure. As a general rule, it is advisable to check properties obtained by
calculation by physical testing so that the analytical model can be refined and made more
relevant to the actual structure.
This Appendix deals with calculation of modal properties and aims to point out some
of the more common sources of error. In doing this, it extends and replaces the Advisory
note on calculation of natural frequencies of grandstand seating decks published in The
Structural Engineer, Vol. 81, No.22, November 2003. The need to check calculations by
testing grandstands is dealt with in Section 5 of the Recommendations whilst Appendix A4
provides advice on specification, procurement and reporting.

A3.2 Modal analysis and natural frequencies


When a linear elastic structure vibrates under arbitrary loading, it does so in a way that
its deflection shape at every moment in time can be presented as a linear combination
of deformed shapes called mode shapes (Clough and Penzien, 1993). Therefore, for a
system with N degrees of freedom, the vector of N unknown displacement functions xi(t)
can be expressed as:
N
" x] t g, = / q ] t g " z , 
r r (A3.1)
r=1

where {zr} is the rth mode shape and qr(t) is the rth time-dependent scaling factor.

Dynamic performance requirements for permanent grandstands subject to crowd action 33


In other words, when calculating the unknown response, xi(t), equation A3.1 can
be written as:
N

x i ] t g = / zir q r ] t g
 (A3.2)
r=1

where zir is the mode shape amplitude of the rth mode at the point and in the direction
of the displacement xi.
The mode shapes {zr} are properties which depend only on the mass and stiffness of
the structure and do not depend on the dynamic loading, so they do not change with time.
The scaling factors qr(t), also known as generalised or modal coordinates, are
functions of time and depend on the dynamic loading. The generalised coordinates qr(t)
are solutions of the following modal equations of motion:

m r pq r ] t g + c r oq r ] t g + k r q r ] t g = Fr ] t g r = 1, N  (A3.3)

where mr, is modal mass for the rth mode, cr is modal damping, kr is modal stiffness
and Fr(t) is modal force given by:
i=N

Fr ] t g = /z ir if ] t g (A3.4)
i=1

and where fi (t) is the time-varying physical force acting at the point and in the
direction of the displacement xi.
For many practical systems in which mass is modelled using only translationally
moving lumped masses, mi the modal mass mr can be calculated as:
N

mr = /m z i
2
ir  (A3.5)
i=1

Equation (A3.3) will be recognised as a group of r independent equations. Each describes


a single degree of freedom system having its own distinct displacement mode. In these
terms, each equation requires its own modal properties: mr, cr and kr associated with the rth
mode. An equation can be set up for each of N modes of vibration. Therefore, there will be
N natural frequencies ωr:
kr
mr  (A3.6)
1
~r = 2 r

Natural frequencies and the corresponding mode shapes are properties of the structure. All
flexible structures have natural frequencies.
Equation (A3.6) shows that the natural frequency ωr that relates to a particular mode
of vibration depends on terms contributing to the modal stiffness kr of the structure and
terms contributing to the modal mass mr of the material that moves during vibration in a
particular mode.
From Equations (A3.3) and (A3.4), it follows that if the external harmonic force
fi (t) has a frequency identical or close to the natural frequency ωs there will be a strong
resonant response qs(t) of mode s. Equation (A3.1) then suggests that this would cause the
physical response to be dominated by mode s.
Equation (A3.4) also suggests that a mode of vibration will be excited by an external
force only if this acts at a point where the amplitude of the mode shape is non-zero.

34 Dynamic performance requirements for permanent grandstands subject to crowd action


A3.3 Basic errors: two prime suspects
A3.3.1 Distinction between force and mass
Equations (A3.5) and (A3.6) make clear that natural frequency involves mass and
not force.
Accordingly, the input to a dynamic analysis for natural frequency should include all
the mass that would be involved in free vibration. This will include the mass of the basic
structure plus the mass of all the attachments including seating, partition walls or barriers,
hospitality boxes and equipment supported by the seating deck.
Note that, in a static analysis to determine forces or bending moments used in strength
calculations, the weight of some of the above elements would be treated as forces acting
on the primary structure rather than as masses that affect dynamic behaviour.
It follows that it is not generally possible to use the analytical model used in a static
analysis by merely adding the dynamic loading. The model will need to be reassessed to
include lumped masses associated with self weight of both the structure and its appendages.
Almost invariably, forces used in static analysis that derive from weight will correspond to
mass to be considered in a dynamic analysis.
It is easy to overlook the contribution of mass due to a non-structural element or,
in the as-built structure, additions made during fit-out. If there are significant differences
between calculated results for natural frequency and those obtained from testing, it is
suggested that one of the first tasks in checking should be a ‘mass audit’ to see if all the
mass that moves is included in the analytical model.

A3.3.2 Stiffness
The stiffness of so-called non-structural elements is often ignored in a static analysis for
forces and bending moments. Errors result if this is done in a dynamic analysis for natural
frequency. In particular:
• The stiffness of a seating deck may be underestimated if the stiffness of wing walls,
vomitories, partitions and glazing attached to the deck are ignored.
• The stiffness of a seating deck may be overestimated if a rigid or fixed boundary
condition is assumed at the interface with the remainder of the structure.
• The stiffness of a cantilever deck is very much dependent on the boundary
conditions of the cantilever. Therefore, it depends on the stiffness of the connection
with the structure supporting the cantilever and by the displacements/rotations in the
supporting structure itself at the location of the connection. Clearly, if the access or
main supporting structure for a cantilever grandstand is itself flexible, the motion of
the cantilever could be determined as much by the motion of the support structure
as by the flexibility of the cantilever. Here the need is to check that the analytical
model is sufficiently extensive to include the influence of the supporting structure.
It will be recognised also that whole-body movement of the support structure can
occur due to foundation movement or flexible tie-backs and may not be solely
determined by the elements of the main structure.

It will be noted also that the mode shapes obtained from testing can be particularly
informative in guiding revisions of an incomplete analytical model.

Dynamic performance requirements for permanent grandstands subject to crowd action 35


A.3.4 Methods for calculation of natural frequencies
A3.4.1 General comment
All methods for calculating natural frequencies fit into one of two groups: the approximate
method suitable for hand calculations (Blevins, 1995), and numerical methods, suitable for
computer applications (NAFEMS, 1992).
The two methods are complementary and should be used in parallel as a cross-check
whenever possible. Rules of thumb provide a useful check rather than a reliable tool for
analysis. Even for simple structures, such rules are unlikely to be sufficiently accurate for
assessment or design but can provide a useful check for gross errors in natural frequencies
obtained by numerical, typically finite element, analysis or other more involved approximate
methods.

A3.4.2 Approximate analysis


These methods typically yield the natural frequency of the fundamental mode of vibration
and are based on treating this mode as a single degree of freedom system having an
assumed, rather than calculated, mode shape. Usually the mode shape is assumed to be
identical to the static deflection profile due to self weight applied in the direction of the
required mode (vertical or horizontal).
The approximation can work quite well for simple single-span or cantilever structures.
More complicated structures require a full numerical analysis because of difficulties in
estimating the relevant mode shape.
Typically, the lowest natural frequency in vibration cycles per second (i.e. in Hz)
can be approximated by a non-dimensional constant A divided by the square root of the
deflection under dead loading, D, i.e.
A
f1 =  (A3.7)
D

For simple beam or cantilever structures, the ‘rule of thumb’ is that A usually lies between
15 and 20 when D is expressed in millimetres. More complex structures may have an A
value outside this range.

A3.4.3 Computer based analysis


This is typically done using standard commercial computer software based on finite element
analysis. Natural frequencies are found as a result of the so called modal analysis or eigen-
value extraction resulting in mode shapes and the corresponding natural frequencies.
For stands comprising nominally identical frames, a 2D ‘plane frame’ or a 3D ‘slice
frame’, finite element analysis can be performed taking into account contributory mass
and stiffness corresponding to the frame. If the frames making the stand are not nominally
identical (e.g. they change width or height, or have generally different mass and stiffness
properties) a full 3D multi-frame analysis is recommended.
Uncertain modelling parameters (e.g. boundary conditions, composite action,
material properties, and effects of cracking) should be studied parametrically to explore
the uncertainty in the calculated values of dynamic properties. If wide variations occur in
the calculated values, the analytical model may need to be revised.

A3.5 Consequences of mistaken idealisations


It is clear that the engineer should be aware of the effects of the various assumptions made
in calculating natural frequencies and, in analysis, should attempt to represent as closely as
possible the geometry of the entire structure and the effect of the non-structural elements.

36 Dynamic performance requirements for permanent grandstands subject to crowd action


In general, assumptions should be made so that the calculated natural frequency will be on
the low side, so providing a conservative estimate for assessment purposes.
Mistaken idealisations may have the effects shown in Table A3.1 on calculated
values of natural frequency.

Table A3.1: Idealisation and its effects on the calculated natural frequency
Idealisation Calculated natural frequency
Neglect of significant mass including neglect of mass Too high
of non-structural elements
Connections taken as rigid when flexible Too high
Only part of the structure considered with the Usually too high: with prospect
remainder taken as motionless and effectively rigid of missing a significant mode
Stiffening effects of the ‘not considered part’ of the Too low
structure more important than the effects of its mass
Wrong/inappropriate mode shape used in Too high
approximate method
Neglect of foundation flexibility Too high
Use of too coarse a finite element grid/mesh Potentially too high
Concrete assumed to be uncracked Too high
Neglect of stiffness arising from interaction between Too low
rakers and seating deck
Assume perfect supports and connections and Too high
neglect of physical slack due to tolerance
Neglect of stiffness of non-structural elements Too low

A3.6 Comment
It should be recognised that dynamic analysis to obtain modal properties of a real structure,
or to estimate its performance under dynamic loading, is a more challenging task than
checking for strength under essentially static loading.
In calculations for strength, assumptions can be made on structural behaviour
concerning connections, ductility and load redistribution that can be made real in design
through appropriate detailing. Also, high accuracy is not essential provided the assumptions
made are conservative and the resulting structure has a reserve of strength over that called
for in the specification. This is not the situation with dynamic analysis. The structure
behaves linearly and as it is constructed. There are no opportunities for alternative load
paths and the simplifications of behaviour based on ductility, that provide safe routes to
simpler design when considering strength under static loading, are not available.
Experienced engineers recognise these difficulties and how they are compounded
by the task of creating an analytical model that will predict adequately the performance
of a ‘still to be constructed’ grandstand. The analytical process itself is almost routine.
However, the modelling is often likely to be less than precise and open to surprise for all
but the simplest structures. In these circumstances, engineers should normally welcome the
opportunity for an independent check of the modal properties by testing and the additional
insight this can provide. (See Appendix 4).

Dynamic performance requirements for permanent grandstands subject to crowd action 37


A3.7 Further information
NAFEMS. A finite element dynamics primer. Glasgow: National Agency for Finite Element
Methods & Standards, 1992.
Clough, R.W. and Penzien, J. Dynamics of Structures. 2nd ed. New York: McGraw-Hill,
1993.
Blevins, R.D. Formulas for natural frequency and mode shape, Malabar, FL: Robert E
Kreiger Publishing Company, 1995.

38 Dynamic performance requirements for permanent grandstands subject to crowd action


Appendix 4  Dynamic testing of
grandstands and seating decks

A4.1 Introduction
The Recommendations refer to the need to determine or check values of natural frequencies
and other modal properties by testing grandstands as fitted out for use, but empty of people.
(See Sections 4 and 5 dealing with dynamic properties and testing).
The Recommendations put particular reliance on Listed Engineers to advise on the
form of testing required. (See Listed Engineers, Section 3.4). This Appendix provides
additional guidance to the Listed Engineer and also to Management, who may require
testing as part of the hand-over procedures for new structures, and for Local Authority
Engineers with responsibilities for Building Control and Safety Certification. The
Appendix extends and replaces the Advisory Note, Dynamic testing of grandstands and
seating decks published by the Institution of Structural Engineers in 2002.
No guidance is given on monitoring grandstands in service. References to monitoring
are given in the Bibliography, (Appendix 5).

A4.2 What should be tested and what results are needed?


The form and extent of testing needed will depend on choice of method – Route 1 or
Route 2 – used in design or assessment.
The Route 1 method needs only the value of the relevant natural frequency for
vertical excitation by crowd action to be determined. (See Section 4.4). Here testing must
be concerned with identifying and determining the lowest natural frequency at which
people can excite the seating deck and feel its motion. For simple structures with easily
identified modes of vibration relatively simple test methods can be used.
If Route 2 (based on performance calculations and satisfying a response based
criterion) is used, all the frequencies that are judged to contribute significantly to the
dynamic response need to be measured so that the dynamic model of the grandstand can
be made as accurate as possible. This requires greater coverage from the test programme
with initial calculations from the theoretical model being used to indicate which modes are
important and therefore which frequencies need to be confirmed experimentally.
Broadly, two types of test are available, corresponding to different levels of
information obtained from the tests.
• Type 1 Tests provide basic information concerning the minimum relevant natural
frequency.
• Type 2 Tests provide more detailed information than Type 1 tests that, in principle,
could provide a full modal description comprising natural frequencies, mode shapes,
modal damping ratios and modal masses for all the modes that are considered to be
important in the dynamic performance calculations. Tests of this type may be required
to check the values of modal properties used in a Route 2 analysis of performance or
when the results of Type 1 testing cannot be reconciled with calculated values.

The Listed Engineer will need to specify the type of testing required and which modes
of vibration need to be investigated. In addition, it will normally be useful if the testing
provides additional information on mode shapes to assist in assessing the significance of
any differences between test and calculated results in a Route 2 approach and also the
degree to which any mode is likely to be excited by crowd movement.

Dynamic performance requirements for permanent grandstands subject to crowd action 39


The Listed Engineer should also consider where additional information may be
needed. This could occur if the Management had set particular performance specifications
for parts of the structure or if information were needed in the context of a possible upgrading
of the structure or for introducing damping systems to reduce the effects of vibration.
Type 1 testing will normally be sufficient to meet the requirements of the
Route 1 approach to satisfying the Recommendations and so enable the Listed Engineer to
recommend a category of use. Moreover, if Type 1 testing is conducted in such a way that
mode shapes are determined in addition to values of natural frequency, comparisons can
be made with the calculated mode shapes as a check that natural frequencies obtained by
test and calculation are being compared on a like-for-like basis.
Type 2 testing will normally be required if a performance based analysis using the
Route 2 approach is undertaken. Such testing will allow the designer to assess the accuracy
of the natural frequencies and also the mode shapes in the theoretical model. Testing will
also allow the experimentally determined modal damping values to be used directly in the
performance calculations since damping cannot be determined theoretically.
Type 2 testing requires more time on site, more specialist equipment and more
information processing than Type 1 testing. As a consequence, use of Type 2 testing is
likely to cost more than a basic Type 1 test programme. In deciding the form of testing
to be used, the additional cost of Type 2 testing needs to be considered in relation to
the increased detail and quality of information that can be obtained and the consequent
increased surety in achievement of safety and better informed management procedures.
To summarise, a Route 1 approach will most often use a Type 1 test with the aim of
determining the lowest relevant natural frequency whereas a Route 2 approach will require
a Type 2 test and aim to determine the natural frequencies, likely to be significant in a
dynamic response calculation, together with their associated modal properties. It should
be noted though that, for some structures, it may not always be possible to identify the
‘relevant’ natural frequency using only Type 1 testing.

A4.3 Analysis and testing


Whether following a Route 1 or 2 approach, the Listed Engineer will be concerned that the
values of the relevant natural frequencies have been determined with sufficient accuracy
for an appropriate Design Event Scenario to be selected on the basis of natural frequency
alone or for a performance based analysis to be relevant to behaviour of the as-built
structure. In doing this, and bearing in mind the idealisations made in even a sophisticated
analysis, it should be realised that exact correspondence between measured and calculated
values is extremely unlikely. However, if the difference in results is substantial, the Listed
Engineer could be expected to review the results to check inter alia that:
• the mode shapes corresponding to the relevant natural frequencies are the same for
the calculated results as those found from physical testing,
• the mass and stiffness of the structure, and all the non-structural elements associated
with the grandstand or seating deck, has been properly represented in the dynamic
analysis, and,
• the support conditions, including the continuity and fixity of the elements of the
structure, are appropriately represented in the dynamic analysis.

Such a review could indicate whether the test programme had missed the mode of vibration
corresponding to the minimum natural frequency in a Route 1 approach or a relevant mode
in a Route 2 approach. The review could show whether the analysis should be refined to

40 Dynamic performance requirements for permanent grandstands subject to crowd action


include some mass or stiffness that had been ignored in an earlier calculation or extended
to examine the significance of assumptions made in modelling the structure. Depending on
the circumstances, the Listed Engineer might decide that further Type 1 testing is necessary
or that more detailed testing is required and so review the testing specification to provide
for some form of Type 2 testing.
It is recommended that an analysis to give an estimate of natural frequencies and
mode shapes should be undertaken before testing is commissioned or undertaken. With
the Route 2 approach, preliminary dynamic response calculations should be performed
to indicate which modes are likely to be significant and the sensitivity to errors on the
frequencies could be explored. In doing this, it will be important to check that the available
drawings properly represent the existing structure with any differences being noted for
future reference. Only by having the results of an analysis available can additional tests
be requested to find a missing mode within a single programme of testing, so avoiding the
Test Agency having to make a second visit to site. Information on the likely mode shape is
also helpful in informing the choice of test points to be used in the test programme.
As noted in the Recommendations, (See Section 4.4 on ‘relevant’ natural frequency
and Section 5.2 on Aims of Testing) it is important to recognise the existence, for some
stadia, of so called ‘global modes’ of vibration. These are typically low frequency modes
involving motion of the entire grandstand structure in sideways sway, twist or front to back
movement causing ‘nodding’ of the seating decks. Care needs to be taken to identify global
modes in the test process, particularly since some excitation methods (e.g. heel‑drop) are
unlikely to provide the energy required to excite such modes.

A4.4 Principles of dynamic testing


A variety of excitation methods may be used in dynamic tests on grandstands. These include
the use of ambient excitation due to wind or other disturbance, impact testing by heel‑drop
or calibrated hammer or the use of shakers to provide excitation at a given frequency.
Whichever method of testing is used, it is vitally important that the level of excitation
available is sufficient to excite all the modes of interest. Also, the instrumentation must
be appropriate to record the response of the grandstand with sufficient accuracy to enable
meaningful results to be derived.
Different testing procedures may be adopted depending on the type of excitation
being used, the availability of instrumentation and the amount of detail required in the
results. For instance, testing may be undertaken with accelerometers at a number of fixed
locations with the excitation sources being moved, from test point to test point, in order to
excite different modes and explore the sensitivity of the structure to vibration. Alternatively,
the excitation source can be used in one place and the measuring points changed from test
to test. A combination of these two approaches may also be appropriate. Also, if mode
shapes are to be determined, testing should be performed across a range of test points that
are sufficiently closely spaced that mode shapes are uniquely defined, even when there are
neighbouring modes with similar shape.
In contrast to ambient vibration surveys, heel-drop tests and some forms of measured
impact testing, shakers provide a consistent and reproducible source of excitation. This
means that, besides providing good quality results for Type 1 testing, shakers can be used
for Type 2 testing with the scope, or range of results, being dependent on the experience
of the operators and the particular techniques and instrumentation employed. As with all
forms of excitation, it is important that the shaker provides sufficient energy to excite the
structure at the frequencies of interest.
An overview of the different techniques is given in Table A4.1. A more detailed
assessment of the different excitation options is given in the following section.

Dynamic performance requirements for permanent grandstands subject to crowd action 41


Table A4.1. Techniques for dynamic testing of grandstands and seating decks
Test characteristics Test outcomes

42
Essential outcome Desirable outcome Additional information relevant to Route 2 analysis
Test Type Excitation Force Natural frequencies Mode shapes Damping ratio Frequency Modal mass
measurement Response Function
Type 1 Ambient Not possible. Yes, but care needed Yes, if excitation Not reliable No No
with interpretation. energy is sufficient.
Can combine
with other Type 1
techniques to assist
interpretation.
Type 1 Heel‑drop Not normally Suitable for simple Provides coarse Not reliable No No
done. structures. Difficulties indication
with complex sufficient for
structures or closely simple structural
separated vibration arrangements.
modes. Not suitable
for global modes.
Type 1 Drop-weight Measured. Yes Yes Better than Possible with further processing
or sledge heel‑drop of measured data if excitation
hammer energy is adequate.
Type 1 or 2 Shaker with Measured Yes, and provides Yes Yes Quality of results dependent on
according to variety of or inferred better quality results technique and instrumentation.
techniques and possible depending on than heel-drop, Most reliable results obtained with
instrumentation types and technique. impact or AVS. instrumented shaker giving direct
employed techniques measurement of force time history
and using multiple-degree of
freedom curve fitting procedures.
Note
The Table provides an initial guide to the choice of test method. Section A4.5 provides more detailed information on the use of the different
methods. However, it is important to discuss with a prospective Test Agency the methods that might be appropriate for a particular situation, how
these would be implemented and the type and quality of results that the particular Test Agency can provide for a given method and programme
of testing.

Dynamic performance requirements for permanent grandstands subject to crowd action


A4.5 Excitation sources and testing techniques
A.4.5.1 Ambient vibration survey (AVS)
This method relies on ambient excitation – typically wind or passing traffic – to excite the
structure. The response is measured and spectra are calculated to yield vibration property
estimates, either by visual inspection of the spectra or by some form of curve fitting. The
method can yield useful results for natural frequencies and mode shapes, if the ambient
excitation is able to excite the modes of interest adequately, but results for modal damping
can be unreliable.
Care is needed in interpreting the results of an ambient vibration survey so that the
relevant modes are correctly identified. (See Recommendations, Section 4.4 Relevant Natural
Frequency). For example, it would be misguided to focus on ‘roof modes’, which engage
only slightly with the seating deck, rather than those modes that engage the seating deck
strongly and so could be excited significantly by crowd loading. Also, care must be taken to
avoid misinterpreting peaks in the AVS response spectra that are not primarily due to resonant
response but correspond to dominant frequencies in the ambient excitation spectrum such
as might occur due to vortex shedding or the influence of machinery in the vicinity. This
is because the method is based upon the excitation spectrum being flat, or at least smooth,
across the frequency range of interest. Also, as the AVS approach depends on the ability
of the ambient excitation to vibrate the empty stand, its application may be limited when
measuring vertical modes of seating decks that are protected from the wind.
Bearing these reservations in mind, AVS should be regarded as a Type 1 test. The
test can be used in isolation but, because of the potential difficulties of interpretation, it
is normally better used in combination with other techniques. However, because of the
difficulties of providing horizontal excitation with significant energy input, it is useful to
recognise that an ambient vibration survey may be the only practicable method of detecting
the presence of global horizontal modes of grandstand vibration.

A.4.5.2 Heel-drop testing


This method is suitable for Type 1 testing of moderate size, simple structural arrangements.
Heel-drops should be performed across a grid of test points making sure that all relevant
modes of vibration are excited. The natural frequencies of excited modes of vibration will
usually show as peaks in the spectra in the response to the heel‑drop.
The signal received from a heel‑drop test usually contains significant noise which
overlays the primary effect of the impact. This leads to a poor signal-to-noise ratio and
spikes in the spectra that may obscure the peaks corresponding to the modes of vibration.
As a consequence, there are difficulties in identifying peaks relating to different modes if
these occur at frequencies that do not differ very much in value.
It is possible to use one or more heel‑drop tests to establish mode shapes by measuring
the response across a series of test points. However, if results from several heel‑drops are
combined, the results may be too crude to enable useful comparison to be made with
results from analysis because of variations in excitation between heel‑drops. It is unlikely
that heel‑drop testing by an individual will excite ‘global’ modes involving significant
motion of the whole of a large stand.

A.4.5.3 Measured impact testing


Simultaneous measurement of an impact force pulse and the corresponding structural
response would enable the full set of modal properties to be determined. The technique is
commonly used in laboratory testing and testing of smaller structures, but does not appear
to have been used on grandstands where the energy required to excite the structure is much
greater. It is possible that, for large structures, an instrumented sledge-hammer or a drop

Dynamic performance requirements for permanent grandstands subject to crowd action 43


weight and force plate could be used. However, as for a heel‑drop, sledge-hammers may
not be sufficiently large to excite the structure sufficiently and drop weight devices may be
difficult to install on a grandstand and could damage the seating deck.
Bearing these difficulties in mind, it is considered that measured impact testing
should be considered as a Type 1 test being, in effect, an upgraded heel‑drop test in which
the excitation is both measured and more repeatable.

A4.5.4 Shaker testing of different types and complexity


There is a wide variety of equipment and techniques available using shakers for dynamic
testing. As might be expected, these have been developed furthest in the context of
mechanical and aerospace engineering. The latest techniques are now becoming more
widely used for site testing structures of significant size so that, from a structural engineering
viewpoint, it can be anticipated that the near future will bring greater choice in the type of
testing that can be employed and the quality of results that can be obtained.
In the past, most dynamic testing of grandstands using shaker excitation has used
rotating mass shakers. Typically, a constant speed of rotation is used to develop sinusoidal
excitation at a particular frequency. The tests are normally repeated for a range of
increasing speeds, so providing excitation at a range of discrete frequencies, a procedure
referred to as stepped sine excitation. The rotating mass shakers that have been used for
grandstands are sufficiently large (i.e. the rotating mass produces sufficient force) to excite
a cantilever deck for all the modes of interest. The amplitude of the excitation force is
easily calculated knowing the rotating mass, its eccentricity and the speed of rotation.
Together with measurements of acceleration from locations on the structure, this enables
natural frequencies, mode shapes, modal mass and damping values to be estimated for
well-separated modes where single degree of curve fitting is appropriate. This capability
goes significantly beyond what is needed for Type 1 testing.
As yet, it has not been the practice to instrument the rotating mass shaker so as to
record directly the excitation force/time history required for correct estimation of modal
properties when there are closely spaced modes of vibration that often occur in cantilevered
seating decks. However, some Test Agencies have developed procedures to derive the phase
difference between the excitation and response and so improve the identification of modes
corresponding to closely-spaced natural frequencies. This allows the full range of modal
properties to be determined and, in these terms, tests using a non-instrumented rotating
mass shaker meet all the requirements for Type 2 testing. However, the processing of the
results involves curve fitting to establish the modal parameters. The accuracy of the results
of curve fitting depends on the quantity and quality of information available to define the
relationships being described. Accordingly, there will be circumstances when the results
of Type 2 tests undertaken with a rotating mass shaker, without additional instrumentation,
can be less accurate than if the force/time history had been obtained by direct measurement
and the results used when processing the data to determine modal properties.
Testing with fully instrumented shakers providing a direct measurement of the
force/time history has been standard practice for some time in mechanical and aerospace
engineering and is now being used for large–scale civil structural engineering applications.
This has led to the use of electrical or hydraulic shakers that provide excitation using
an inertial mass oscillating in the direction of excitation. These tend to be smaller than
rotating mass shakers but, being portable, can be moved around a structure to provide
excitation at different locations.
Simultaneous measurement of the shaker excitation force and the corresponding
response is used when estimating the Frequency Response Function (FRF) between the

44 Dynamic performance requirements for permanent grandstands subject to crowd action


response and excitation on a test structure. In this case, both modulus and phase information
for the FRF is determined directly and used in a curve fitting approach that yields the
natural frequency, mode shape, modal damping ratio and modal mass for all the modes
of interest. The availability of a complete FRF makes this method the most reliable of all
those available, particularly where modes are closely spaced in frequency. Stepped-sine
and slow-sweep sine as well as broadband random excitation can be used, depending on
the time available for testing, the shaker employed (typically inertial, with acceleration of
the mass measured so as to derive the force) and the facilities for information processing.
Always providing that the shaker generates sufficient force to excite all the modes of
interest, instrumented shakers provide high quality information for both Type 1 and
Type 2 testing.
Because of the substantial size of a grandstand seating deck, not all commercial
shakers will be suitable for testing grandstands. More particularly, the shaker has to be of
sufficient size that the oscillatory or rotating mass develops the force necessary to excite
all relevant modes of the structure. In practice, this difficulty is avoided by using more
than one shaker to achieve the necessary excitation. Besides increasing and distributing
the energy input, simultaneous use of shakers at more than one location helps to improve
the identification of modes with closely spaced natural frequencies and to minimise the
possibility of missing a mode of vibration as could occur with a shaker used in a single
location.

A.4.5.5 Future developments


It will be evident that there are techniques available in other disciplines that, if properly
implemented, can enhance the general capability for dynamic testing of grandstand
structures. Almost certainly, greater choice of testing techniques for use on grandstands
will become available to the Listed Engineer responsible for procuring testing and using
the results. In making this choice, the Listed Engineer will need to consult the proposed
Test Agency to ascertain its range of expertise and preferred way of working. In a climate
of change and new developments, an established track record of on-site dynamic testing of
structures of significant size will be a useful recommendation.

A4.6 Specification and procurement


The role of the Listed Engineer in preparing the specification, procuring testing and
reporting to management is illustrated in Figure A4.1. Note that the need for Type 2
testing may emerge from the procedures indicated if additional information is considered
necessary in order to reach a satisfactory conclusion.
Prior to commissioning a Test Agency to undertake work, the Listed Engineer
should decide whether the Route 1 or 2 approach is to be followed, the extent and types
of information required from testing and take a preliminary view of the techniques to be
employed. At this stage, it will often be useful to discuss the programme and methods
with a possible Test Agency and agree a specification for the work required. The Test
Agency should have the necessary experience and capability to undertake the chosen type
of testing and deal with the logistical difficulties of on-site testing of major structures. It
should also be noted that, although Type 1 tests are simpler than Type 2, there is the same
need for experienced personnel to undertake the testing and reporting. For instance, the
equipment required for heel-drop tests or ambient vibration monitoring is quite widely
available but can be used by inexpert operators so giving results of little value. It is also
useful if the Test Agency is able to process and analyse results, at least partially, on site
so allowing some flexibility in the programme and avoiding the need for repeat testing on
another occasion.

Dynamic performance requirements for permanent grandstands subject to crowd action 45


Start In the particular circumstances of the
structure under consideration, is testing No
required to confirm the assumptions
made in design or assessment?
(See Section 3.4)

Yes

Has dynamic testing been Calculate relevant natural frequencies.


No previously undertaken on the
stand in its present form and
is the report of the testing
available?

Does the comparison give confidence


Yes that the figures are sufficiently accurate
to support a recommendation
confirming the assumptions made in
Compare test and design or assessment?
calculated values.

No Yes

Prepare report for


Decide information management
required from testing or
additional testing. (See Appendix 4
Section 4.7)

Agree method and End


extent of testing with
Test Agency.

Write requirement
specification and
recommend experienced
Test Agency to Management.

Testing undertaken by Test Review test results and compare


Agency. The Listed Engineer with calculated values. Refine
should be available to review the analytical model if necessary
results and revise programme and compare test and
if necessary during testing. recalculated results.

Figure A4.1  The role of the Listed Engineer

46 Dynamic performance requirements for permanent grandstands subject to crowd action


The specification should be written as a ‘performance’ document, based on the
agreed types of testing and the properties required from the tests, with the details of the
test programme being left to the Test Agency to decide. However, the specification should
include:
• A description of the properties required from testing.
• Confirmation of the agreed type of testing and the form of presentation of results.
• A requirement for the work to be undertaken to a recognised quality standard such
as BS ISO Standard 14964:2000, ‘Mechanical vibration and shock – Vibration
of stationary structures – Specific requirements for quality management in
measurement and evaluation of vibration’.
• The time agreed for the delivery of the results and report on the testing.
• Requirements for reporting, bearing in mind that the Engineer’s report to
Management should include ‘an account of the procedures used and the detailed
results’.
• A requirement for a method statement to meet the requirements of Health and
Safety, and also CDM Regulations where these are applicable.

The Listed Engineer should be available while testing is in progress in order to review
results as they are obtained and, if necessary and possible to arrange, modify the instructions
to the Test Agency

A4.7 Reporting
The Listed Engineer is required to make a report to Management. This should include:
• An explanation of the use of personal judgement in requiring additional testing in
situations where test records are already available or of a decision not to test (as
outlined in Section 3.4).
• A note on the choice of Test Agency including the Agency’s track record of on-site
testing of structures.
• The agreed specification for the test programme.
• The Test Agency’s report of the testing including all results.
• An interpretive appraisal of the results including comparisons with values used in
design and the significance of any differences in estimates of performance.
• Recommendations on any further action required.

A4.8. Further Information


More information on methods of testing may be obtained from the following sources:
Dynamic Test Agency. Primer on best practice in dynamic testing. London: Chameleon
Press, 1993.
Maia, N.M. and Silva, J.M.M. eds. Theoretical and experimental modal analysis. Baldock:
Research Studies Press, 1997.
Ewins, D.J. Modal testing: theory, practice and application. 2nd ed. Baldock: Research
Studies Press, 2000.

Dynamic performance requirements for permanent grandstands subject to crowd action 47


Appendix 5  Bibliography

The Bibliography is a limited selection from published material. Papers are referenced
under the following categories.
A Analytical methods
B Behaviour of grandstands in service
D Loading, dynamic load factors, crowd behaviour and tolerance of motion
G Overviews and general interest
H Human structure interaction
T Testing and monitoring of grandstands
M Management, risk assessment and liability
S Codes, Standards and Guidance.
∗ Referred to in other Appendices

The entries are given in date order by year. References to papers of particular interest or
relevance are annotated with a comment in italics.

G/D Bachmann, H. and Ammann, W. Vibration in structures: induced by man and


machines. Zurich: IABSE, 1987.
Widely used source of information that has stimulated research and practice

S BS 6841:1987: Guide to measurement and evaluation of human exposure to


whole-body mechanical vibration and repeated shock. London: BSI, 1987
G/D Allen, D.E. ‘Vibrations from human activities’, Concrete International: Design
and Construction, 12(6), June 1990, pp66-73.
Wide coverage of Canadian Practice as recommended in successive revisions of Commentaries
to the NBC Building Code

A* NAFEMS. A finite element dynamics primer. Glasgow: National Agency for


Finite Element Methods & Standards, 1992.
A* Clough, R.W. and Penzien, J. Dynamics of structures. 2nd ed. New York: McGraw-
Hill, 1993.
A/T* Dynamic Test Agency. Primer on best practice in dynamic testing. London:
Chameleon Press, 1993.
B Batista, K.C. and Magluta, C. ‘Spectator-induced vibration of Maracana football
stadium’, Proceedings of the 2nd European conference on structural dynamics:
EURODYN ‘93, Trondheim, 21-23 June 1993, vol 2. Rotterdam: Balkema, 1993,
pp985-992.
B Kasperski, M. and Niemann, H.J. ‘Man induced vibration of a stand structure’,
Proceedings of the 2nd European conference on structural dynamics: EURODYN
‘93, Trondheim, 21-23 June 1993, vol 2. Rotterdam: Balkema, 1993, pp977-983.
Full coverage of design issues set in the context of determining remedial measures for a grandstand
that vibrated excessively under crowd loading. Recommends acceleration limits for design and
proposes a 7Hz minimum natural frequency limit for grandstands without additional damping

48 Dynamic performance requirements for permanent grandstands subject to crowd action


B Van Staalduinen, P. and Courage, W. ‘Dynamic loading of Feyenoord Stadium
during pop concerts’, Places of assembly and long-span structures, IABSE
Symposium, Birmingham, 1994, IABSE Reports, vol 71, 1994, pp283-288.
A* Blevins, R.D. Formulas for natural frequency and mode shape. Malabar, FL:
Robert E Kreiger Publishing Company, 1995.
D Ji, T. and Ellis, B.R. ‘Floor vibration induced by dance-type loads: theory’, The
Structural Engineer, 72(3), 1 Feb 1994, pp37-44.
Highly influential paper in deriving DLFs for individual jumping in terms of contact ratio for a
repeated half sine load pulse

D Ellis, B.R. and Ji, T. ‘Floor vibration induced by dance-type loads: verification’,
The Structural Engineer, 72(3), 1 Dec 1994, pp45-50.
Includes experimental confirmation of the half sine load pulse assumption for jumping on a very
stiff support

D Kasperski, M. ‘Actual problems with stand structures due to spectator induced


vibrations’, Proceedings of the 3rd European conference on structural Dynamics:
EURODYN ‘96, Florence, 5-8 June 1996, vol 1. Rotterdam: Balkema, 1996,
pp455-461.
Overview including DLFs for different activities including hand clapping and stamping, review
of tolerance of motion including potential for panic and example of use of tuned mass dampers

S BS 6399-1: 1997: Loadings for buildings. Part 1: Code of practice for dead and
imposed loads. London: BSI, 1996.
First inclusion in UK Code of requirements concerning dynamic loading due to people in
buildings. Ji and Ellis (1996) contact ratios identified with specific activities and so DLFs. Also
set out natural frequency trigger values as alternatives to assessing performance by calculation
A/T* Maia, N.M. and Silva, J.M.M eds. Theoretical and experimental modal analysis.
Baldock: Research Studies Press, 1997.
H Ellis B.R. and Ji, T. ‘Human–structure interaction in vertical vibrations’, Proc.
ICE, Structures and Buildings, 122(1), Feb 1997, pp1-9.
Recognition of role of passive people in moderating motion

S/M Scottish Office and Department of National Heritage. Guide to safety at sports
grounds. 4th ed. London: The Stationery Office, 1997 [the Green Guide].
Mainly concerned with spectator management but includes a brief section on dynamics with
different frequency limits for structures at sports grounds to those given in BS 6399 (1996)

S ISO 2631-1: 1997: Mechanical vibration and shock: evaluation of human exposure
to whole-body vibration. Part 1: General requirements. Geneva: ISO, 1997.
G Reid, W.M., Dickie, J.F., and Wright, J. ‘Stadium structures: are they excited?’
The Structural Engineer, 75(22), 18 Nov 1997, pp383-388.
Useful overview of factors influencing structural design of cantilever grandstands and roofs.
Emphasises difficulty of designing practical grandstands to BS6399(1996) frequency limits

M Chapman, J.C. ‘Collapse of the Ramsgate Walkway’, The Structural Engineer,


76(1), 6 Jan 1998, pp1–10.
Although not concerned with stadia, presents important messages to operators of facilities used
by the public on their responsibilities in procuring services and overseeing the work of specialists

H Wei, L. and Griffin, M.J. ‘Mathematical models for the apparent mass of the
seated human body exposed to vertical vibration’, J. Sound & Vibration, 212(5),
1998, pp855-874.
Major study of passive action leading to body unit properties for seated people

Dynamic performance requirements for permanent grandstands subject to crowd action 49


T Littler, J.D. ‘The dynamic response of a three tiered cantilever grandstand’,
Proceedings of the 4th European conference on structural dynamics: EURODYN
‘99, Prague, 7-10 June 1999, vol 1. Rotterdam: Balkema, 1999, pp623-628.
G/M Standing Committee on Structural Safety. Structural Safety 1997-99: review
and recommendations. 12th report of SCOSS. London: SETO, 1999, Section 3.2:
‘Safety of sports stadia structures’, pp28-29.
Authorative independent expression of concern on procurement, inspections and maintenance of
stadia structures. Also suggests independent checks on structural designs

M Health and Safety Executive. The event safety guide: a guide to health, safety,
and welfare at music and similar events. 2nd ed. Sudbury: HSE Books, 1999.
Key document providing more information than the ‘Green Guide’ for non- sporting events

T Littler, J.D. Permanent cantilever grandstands: dynamic response. BRE


Information Paper IP 5/00. Garston: BRE, 2000.
D Ellis, B.R., Ji, T. and Littler, J.D. ‘The response of grandstands to dynamic crowd
loads’, Proc. ICE, Structures and Buildings, 140(4), Nov 2000, pp355-365.
Notes human structure interaction due to presence of a passive crowd and corresponding effects
on natural frequency of the combined system

A/T* Ewins, D.J. Modal testing: theory, practice and application. 2nd ed. Baldock:
Research Studies Press, 2000.
D Kasperski, M. ‘Safety assessment of stadia in regard to human induced vibrations’,
Safer solutions in sport and leisure: responsibilities for crowd management
at major events, Manchester, 5 April 2001 [unpublished Institution of Civil
Engineers seminar].
Besides loading, and target reliability over life-time use, discusses tolerance of motion and
potential for panic due to excessive motion of a stand

S/D Willford, M. ‘Stadium Dynamics’, Safer solutions in sport and leisure:


responsibilities for crowd management at major events, Manchester, 5 April
2001 [unpublished Institution of Civil Engineers seminar].
Overview including statistical appreciation using Monte Carlo modelling and assumed
distributions of input variables to achieve probabilistic acceleration predictions for different
frequency stands. Concludes deterministic design leads to an overestimate of risk of exceeding
any given acceleration limit due to neglect of song frequency input and other factors
G/M Standing Committee on Structural Safety. Structural Safety 2000-01: 13th report
of SCOSS. London: IStructE, 2001, chapter 3: Dynamic response of structures,
pp23-26.
On cantilever decks at sports grounds, expresses concern on the “consequences of a structural
collapse or disturbing movement causing panic amongst an occupying crowd…”. Concludes
that specifically targeted research needed to resolve uncertainties in design for dynamic crowd
loading
D Ginty, D., Derwent, J.M. and Ji, T. ‘The frequency ranges of dance type loads’,
The Structural Engineer, 79(6), 20 Mar 2001, pp27–31.

50 Dynamic performance requirements for permanent grandstands subject to crowd action


S/M Institution of Structural Engineers, Department for Transport, Local Government
and the Regions and Department for Culture, Media and Sport. Dynamic
performance requirements for permanent grandstands subject to crowd action:
interim guidance on assessment and design. London: IStructE, 2001.
Uses natural frequency for vertical excitation of the stand to set limits on categories of use.
Testing required to check calculated natural frequencies. Knowledge base considered inadequate
to enable recommendations to be formulated for calculating grandstand performance under
crowd loading. Recommendations accepted by UK Government re Building Control and Safety
Certification

D Ellis, B.R. and Ji, T. Loads generated by jumping crowds: experimental


assessment. BRE Information Paper IP 4/02. London: CRC, 2002.
Only study of large group (up to 64 people) loading and reduction in DLFs for jumping with
increasing group size

H Sachse, R., Pavic, A. and Reynolds, P. ‘The influence of a group of humans on


modal properties of a structure’, Proceedings of the 5th European conference on
structural dynamics: EURODYN ‘02, Munich, 2-5 September 2002. Rotterdam:
Balkema, 2002, pp1241-1246.
Wide ranging study providing body unit models for passive human structure interaction effects

D* Littler, J.D. ‘Frequencies of synchronised human loading from jumping and


stamping’, The Structural Engineer, 81(22), 18 Nov 2003, pp27–35.
Reviews song frequencies experienced at concerts and challenges received wisdom on limits to
possible excitation ranges with tests involving jumping and stamping

H Matsumoto, Y. and Griffin, M.J. ‘Mathematical models for the apparent masses
of standing subjects exposed to vertical whole-body vibration’, J. Sound &
Vibration, 260(3), pp431-451.
Passive action and body unit properties for erect people

S/G ISO/CD/10137: 2004: Bases for the design of structures: serviceability of


buildings and pedestrian walkways against vibration [Committee Draft].
Includes proposals for stadia with loading corresponding to an extreme event involving everyone
jumping. Proposes separate acceleration limits for the onset of panic and an upper limit for
comfort. It is doubtful that practical cantilever grandstands would meet these requirements
A/D/S Ellis, B.R. and Ji, T. The response of structures to dynamic crowd loads. BRE
Digest 426. Garston: BRE Bookshop, 2004.
Provides methods for calculating dynamic response in accordance with BS 6399-1 (1996). Useful
guidance for low crowd densities and vigorous activity on relatively stiff structures, (dancing and
aerobics). Not applicable to dense crowd loading on flexible grandstands
D Ellis, B.R. and Littler, J. D. ‘The response of cantilever grandstands to crowd
loads. Part 1: Serviceability evaluation.’ Proc. ICE, Structures and Buildings,
157(SB4), Aug 2004, pp235–241.
Explores the use of the vibration dose value (VDV) approach to assessing tolerance of motion for
people in grandstands

A/D Ellis, B.R. and Littler, J.D. ‘The response of cantilever grandstands to crowd
loads. Part 2: Load estimation’, Proc. ICE, Structures and Buildings, 157(SB5),
Oct 2004, pp297-307.
Back analysis of two cantilever grandstands tiers to determine effective DLFs for crowd loading
at concerts with modest excitation levels. Conventional analysis with allowance for crowd size
(Ellis & Ji 2002) led to a value of 16% effective damping which was attributed to crowd action.
Results applied to other grandstands. High damping clearly the result of human structure
interaction

Dynamic performance requirements for permanent grandstands subject to crowd action 51


H Yao, S., Wright, J. R., Pavic, A. and Reynolds, P. ‘Experimental study of human-
induced dynamic forces due to bouncing on a perceptibly moving structure’,
Canadian J. Civil Engineering, 31(6), Dec 2004, pp1109-1118.
Unambiguous demonstration of effects of human structure interaction due to an active (i.e.
moving) person on a flexible support for a range of natural frequencies and mass ratios typical
of practical cantilever grandstands. Significant reductions in contact force at resonance (drop-
out) observed leading to lower accelerations than would be predicted from conventional theory
ignoring active human structure interaction

T Reynolds, P., Pavic, A. and Ibrahim, Z. ‘A remote monitoring system for


stadium dynamics’, Proc. ICE, Structures and Buildings, 157(SB6), Dec 2004,
pp385-393.
Report of long-term monitoring of moderately flexible grandstand mainly used for viewing
soccer

A/H Sachse, R., Pavic, A., and Reynolds, P. ‘Parametric study of modal properties of
damped two-degree-of-freedom crowd-structure dynamic systems’, J. Sound &
Vibration, 274(3-5), 2004, pp461-480.
T Reynolds, P., Pavic, A. and Willford, M. ‘Prediction and measurement of stadia
dynamic properties’, 23rd International modal analysis conference: IMAC-
XXIII), Orlando, Florida, 31 Jan-3 Feb 2005.
Account of dynamic testing to obtain modal properties of a curved multi-tier cantilever grandstand
with comparison of results from pre test analysis using single frame and 3-D analysis and post
testing results from 3-D Finite Element modelling. The post testing analysis satisfactorily
reproduced the family of closely spaced modes but with discrepancies in values of natural
frequency considered due to the treatment of non-structural elements and omission of the roof in
the FE model

G/H Willford, M. ‘Dynamic performance of stands’, in Culley, P. and Pascoe, J. eds.


Stadium Engineering. London: Thomas Telford, 2005, pp47-54.
Broad coverage of design of stadia for dynamic loading with particular consideration given to
human structure interaction due to passive /inactive people in an otherwise active crowd. Active
participation treated using conventional DLFs
S ‘Commentary D: deflection and vibration criteria for serviceability and fatigue
limit states’, in National Research Council of Canada. User’s guide - NBC 2005:
structural commentaries (Part 4 of Division B). Ottawa: NRC, 2005, ppD1-
D10.
Widely used Code based on performance design principle. 2005 revision includes
recommendations for stadia based on serviceability criteria related to Canadian practice and
“commonly encountered events”. Does not address concerns on possibility of an extreme event
and consequences for safety related to panic due to excessive motion. Not recommended for
major stadia with open-ended use

T Reynolds, P. and Pavic, A. ‘The dynamic performance of sports stadia under


crowd dynamic loading at concert events’, Proceedings of the 6th European
conference on structural dynamics: EURODYN ‘05, Paris, 4-7 September 2005.
Rotterdam: Millpress, 2005, pp473-478.
T/H Reynolds, P. and Pavic, A. ‘Vibration of a large cantilever grandstand during an
international football match’, ASCE J. Performance of Constructed Facilities,
20(3), Aug 2006, pp202-212.
Presents modal properties obtained by testing the empty stand and compares these with properties
obtained whilst monitoring the stand during the match. Significant differences were recorded
showing the crowd interacted structurally with the basic structure

52 Dynamic performance requirements for permanent grandstands subject to crowd action


T Mohanty, P. and Reynolds, P. ‘Modelling of dynamic crowd-structure interactions
in a grandstand during a football match’. 24th International modal analysis

H Yao, S., Wright, J.R., Pavic, A. and Reynolds, P. ‘Experimental study of human-
induced dynamic forces due to jumping on a perceptibly moving structure’,
J. Sound & Vibration, 296, 2006, pp150-165.
Demonstration that active human structure interaction does not depend on uninterrupted contact
with the support with results for jumping similar to those in the 2004 paper by the same authors

D/H Sim, J., Blakeborough, A. and Williams, M. ‘Modelling effects of passive crowds
on grandstand vibration’, Proc. ICE, Structures and Buildings, 159(SB5), Oct
2006, pp261-272.
D* Parkhouse, J.G. and Ewins, D.J. ‘Crowd-induced rhythmic loading’, Proc. ICE,
Structures and Buildings, 159(SB5), Oct 2006, pp247-259.
Results of 1000 tests involving individual bobbing and jumping on force plates at excitation levels
comparable to pop-concert participation and leading to synthesis of Dynamic Load Factors for
groups of different size
H* Dougill, J.W., Wright, J.R., Parkhouse, J.G. and Harrison, R.E. ‘Human structure
interaction during rhythmic bobbing’. The Structural Engineer, 84(22), 21 Nov
2006, pp32–39.
Theoretical development of an active human structure interaction model with properties derived
from independent experiments and validation through comparison with results from individual
bobbing/bouncing on a flexible platform. Includes discussion of relevance to grandstands
H Alexander, N.A. ‘Theoretical treatment of crowd-structure interaction dynamics’,
Proc. ICE, Structures and Buildings, 159(SB6), Dec 2006, pp329-338.
Theoretical treatment of similar model to Dougill et al (2006)

M The Construction (Design and Management) Regulations. Norwich: The


Stationery Office, 2007 (SI 2007/320).
Important implications for a client requiring construction works

T Reynolds, P., Pavic, A. and Carr, J. ‘Experimental dynamic analysis of the


Kingston Communications Stadium’, The Structural Engineer, 85(8), 17 Apr
2007, pp33–39.
Includes comparison of calculated and measured dynamic properties

C Institution of Structural Engineers. Temporary demountable structures: guidance


on procurement, design and use. 3rd ed. London: IStructE, 2007.
Comprehensive guidance on demountable structures. Guidance on horizontal motion is adopted
in the present Recommendations (Table 2) but without allowance for geometrical imperfections

T/B/H* Pavic, A. and Reynolds, P. ‘Experimental verification of novel 3DOF model


of grandstand crowd-structure dynamic interaction’, 26th International modal
analysis conference: IMAC-XXVI, Orlando, Florida, 4-7 Feb 2008, paper 257.
Includes details of testing a stadium before and during a pop-concert with moderate crowd
excitation. Calculations of performance using finite element analysis and the human structure
interaction model recommended in Appendices 1 and 2 show acceptable agreement with the
measured values of RMS acceleration

Dynamic performance requirements for permanent grandstands subject to crowd action 53


54 Dynamic performance requirements for permanent grandstands subject to crowd action
The Institution of Structural Engineers
The Department for Communities and Local Government
The Department for Culture Media and Sport

December 2008

Dynamic performance
requirements for permanent
grandstands subject to
crowd action

Recommendations for management,


design and assessment

The Institution of Structural Engineers


International HQ, 11 Upper Belgrave Street, London SW1X 8BH, United Kingdom
T: +44 (0) 20 7235 4535
F: +44 (0) 20 7235 4294
E: mail@istructe.org
W: www.istructe.org Published by the Institution of Structural Engineers

Stadia covers.indd 2-3 3/12/08 16:13:11

S-ar putea să vă placă și