Springer Tracts in Civil Engineering

Kristian Dahl Hertz

Philip Halding

Sustainable
Light
Concrete
Structures
Springer Tracts in Civil Engineering

Series Editors
Sheng-Hong Chen, School of Water Resources and Hydropower Engineering,
Wuhan University, Wuhan, China
Marco di Prisco, Politecnico di Milano, Milano, Italy
Ioannis Vayas, Institute of Steel Structures, National Technical University of
Athens, Athens, Greece
Springer Tracts in Civil Engineering (STCE) publishes the latest developments in
Civil Engineering - quickly, informally and in top quality. The series scope includes
monographs, professional books, graduate textbooks and edited volumes, as well as
outstanding PhD theses. Its goal is to cover all the main branches of civil engineering,
both theoretical and applied, including:
• Construction and Structural Mechanics
• Building Materials
• Concrete, Steel and Timber Structures
• Geotechnical Engineering
• Earthquake Engineering
• Coastal Engineering; Ocean and Offshore Engineering
• Hydraulics, Hydrology and Water Resources Engineering
• Environmental Engineering and Sustainability
• Structural Health and Monitoring
• Surveying and Geographical Information Systems
• Heating, Ventilation and Air Conditioning (HVAC)
• Transportation and Traffic
• Risk Analysis
• Safety and Security

Indexed by Scopus
To submit a proposal or request further information, please contact:
Pierpaolo Riva at [email protected] (Europe and Americas) Wayne Hu
at [email protected] (China)

More information about this series at http://www.springer.com/series/15088

Kristian Dahl Hertz · Philip Halding

Sustainable Light Concrete

Structures
Kristian Dahl Hertz Philip Halding
Broenshoej, Denmark Department of Civil Engineering
Technical University of Denmark
Kongens Lyngby, Denmark

ISSN 2366-259X ISSN 2366-2603 (electronic)

Springer Tracts in Civil Engineering
ISBN 978-3-030-80499-2 ISBN 978-3-030-80500-5 (eBook)
https://doi.org/10.1007/978-3-030-80500-5

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature
Switzerland AG 2022
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse
of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by similar
or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface

Light concrete structures offer solutions to many of the urgent problems builders and
structural engineers are facing today. The small weight means less consumption of
resources and possibilities of making longer spans. The open microstructure makes
it fire-resistant and creates a sound damping effect. In addition, it is capable of
regulating moisture content of enclosures. You may cast light concrete structures in
textile moulds made with a minimum of materials in many spectacular shapes, since
the pressure on the mould is small.
Super-light structures and sandwich structures use light concrete to fill out the
space between stronger concrete parts, so that the light concrete can stabilize the
strong and protect it from fire. Light concrete has a smaller E-modulus than the
strong, and thereby the engineer can guide forces to follow an optimized path via
the strong concrete parts. This means that the engineer decides where the forces
should be in the structure. This is the opposite of the traditional approach, where the
structure determines what the engineer must do, because the engineer should first
find the path of the forces, before assessing the load-bearing capacity.
Super-light prefabricated SL-deck elements offer a large flexibility regarding
shape and holes for pipes and implemented services as floor heating, sewers, water
and electricity.
The combination of strong and light concrete makes it possible to establish a
sound insulation with a material consumption of only three-fourth of what is elsewise
required.
Prefabricated light floor systems allow beamless solutions with larger spans for
offices, industry and car parks.
You may also utilize the super-light principle with combined application of light
and strong concrete for creating tunnels and especially floating tunnels.
Another specific application is for shipbuilding, where the lightness of the struc-
ture improves manoeuvrability and reduces costs and resource demands compared
to traditional concrete ships.
With pearl-chain technology, you can create curved shapes from simple flat mass-
produced slab elements, which are easy and cheap to cast and transport to the erection
site. This can reintroduce arches and vaults, which elsewise have become too difficult
and costly to build.
vii
viii Preface

Our new research in the lightest concrete qualities of densities down to about fifty
kilogram per cubic metre has developed materials with insulating properties equal
to mineral wool that can be cast out to fit anywhere.
Finally, modern builders and engineers are concerned about the impact on the
climate.
They want to reduce the CO2 emission from material production and building
processes.
The reduced weight of super-light structures means less consumes of cement and
aggregates, and in addition, we can now produce cement with less pollution and
a smaller CO2 impact. These numbers depend on who are producing the materials
and how. However, the durability and long lifetime of the structures make them
sustainable and in many cases more sustainable than, e.g., structures in wood.
It is not new to apply light concrete structures. Roman builders developed remark-
ably skills creating them. Roman concrete structures and technology represent a
source of inspiration for modern builders and engineers. Their ideas and detailing
have proved to be applicable in full scale and just wait for being reintroduced in
combination with modern building technology. Super-light structures can be seen as
examples of that. We therefore introduce the ancient light concrete structures in this
book based on intensive studies on the subject.
The authors have written the present book in order to present the many technolo-
gies and detailed solutions available for light and super-light concrete structures. It
is the intention that the book can serve as a design guide, a reference and a textbook
for teaching the subject. In addition, the book is intended to serve as a source of
inspiration for consulting engineers, architects and others interested in the subject
and in creating a sustainable future.

Broenshoej, Denmark Kristian Dahl Hertz

Kongens Lyngby, Denmark Philip Halding
Acknowledgement

We would like to express our gratitude to the Realdania association for supporting
the writing of the present book.

ix
Contents

1 History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Ancient Building Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1.1 Gypsum and Limestone Mortar . . . . . . . . . . . . . . . . . . . . . 1
1.1.2 Ancient Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Ancient Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.2.1 Walls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.2.2 Arches and Vaults . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.2.3 Cupolas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.2.4 Roman Reinforced Concrete Structures . . . . . . . . . . . . . . 11
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2 Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.1 Heavy Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.1.1 Ordinary Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.1.2 High-Strength Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.2 Light Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.2.1 Light Aggregate Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.2.2 Foam Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.2.3 Aerated Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.2.4 Pervious Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.3 Reinforcement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.3.1 Slack Steel Bars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.3.2 Prestressing Steel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.3.3 Slack Carbon Fibre Reinforcement . . . . . . . . . . . . . . . . . . 31
2.3.4 Prestressed Carbon Fibre Reinforcement . . . . . . . . . . . . . 32
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3 Super-Light Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.1 Super-Light Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.1.1 Minimal Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.1.2 Super-Light Principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
3.1.3 Benefits and Goals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
3.2 Direct Engineering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
xi
xii Contents

3.2.1 Direct Engineering Principle . . . . . . . . . . . . . . . . . . . . . . . 39

3.2.2 Lower Bound Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
4 Slabs and Beams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
4.1 SL-Decks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
4.1.1 Prototypes of Super-Light Deck Elements . . . . . . . . . . . . 45
4.1.2 Mass Produced SL-Decks . . . . . . . . . . . . . . . . . . . . . . . . . . 48
4.1.3 Fire Resistance of SL-Decks . . . . . . . . . . . . . . . . . . . . . . . 54
4.1.4 Sound Insulation of SL-Decks . . . . . . . . . . . . . . . . . . . . . . 54
4.1.5 Sound Damping of SL-Decks . . . . . . . . . . . . . . . . . . . . . . . 55
4.2 Super-Light Deck Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
4.2.1 Fixed End Connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
4.2.2 Connection to Balcony . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.2.3 Connection to Column . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
4.2.4 Safety Handrails . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
4.2.5 Services in SL-Decks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
4.2.6 Blade Connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
4.3 Beamless Super-Light Decks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
4.3.1 Principle of Beamless Decks . . . . . . . . . . . . . . . . . . . . . . . 60
4.3.2 Moment Distribution in Beamless Decks . . . . . . . . . . . . . 62
4.3.3 Design for Disassembly . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
4.4 Light Concrete Slabs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
4.4.1 Massive Light Concrete Slabs . . . . . . . . . . . . . . . . . . . . . . 63
4.4.2 Sandwich Slabs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
4.5 Super-Light Concrete Beams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
4.5.1 Rectangular Super-Light Beams . . . . . . . . . . . . . . . . . . . . 65
4.5.2 Super-Light Beam System for Halls . . . . . . . . . . . . . . . . . 66
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
5 Columns and Walls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
5.1 Light Concrete Columns and Walls with Uniform
Cross-Section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
5.1.1 Centrally Loaded Light Concrete Columns
and Walls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
5.1.2 Eccentrically Loaded Uncracked Columns
and Walls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
5.1.3 Columns and Walls with Plastic Compression . . . . . . . . . 77
5.1.4 Beam-Columns and Cracked Walls . . . . . . . . . . . . . . . . . . 78
5.1.5 Design of Columns and Walls—Summary . . . . . . . . . . . . 81
5.2 Examples Columns and Walls with Uniform Cross-Section . . . . . 83
5.2.1 Example of a Tested Column . . . . . . . . . . . . . . . . . . . . . . . 83
5.2.2 Super-Light Wall Design . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
5.2.3 Example Sandwich Wall . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
5.2.4 Example Wall and Column with Circular Light
Cores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Contents xiii

5.3 Light Concrete Columns and Walls with Entasis . . . . . . . . . . . . . . 94

5.3.1 Centrally Loaded Columns and Walls with Entasis . . . . . 94
5.3.2 Example of a Column with Entasis . . . . . . . . . . . . . . . . . . 97
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
6 Pearl-Chain Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
6.1 Pearl-Chains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
6.1.1 Pearl-Chain Principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
6.1.2 Prestressing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
6.1.3 Pearl-Chain Elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
6.2 Pearl-Chain Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
6.2.1 Pearl-Chain Arches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
6.2.2 Integrated Pearl-Chains . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
6.2.3 Cross-Stringer Structures . . . . . . . . . . . . . . . . . . . . . . . . . . 110
6.2.4 Integrated Edge-Beams . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
6.2.5 Tension Ties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
6.2.6 Drilled Tension Ties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
7 Arch Bridges and Vaults . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
7.1 Arches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
7.1.1 Arch Principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
7.1.2 Arch Shapes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
7.2 Pearl-Chain Arches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
7.2.1 Pearl-Chain Arch Bridges . . . . . . . . . . . . . . . . . . . . . . . . . . 122
7.2.2 Long Span Pearl-Chain Arch Bridges . . . . . . . . . . . . . . . . 123
7.2.3 Half-Arch Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
7.3 Pearl-Chain Sandwich Arches and Vaults . . . . . . . . . . . . . . . . . . . . 127
7.3.1 Pearl-Chain Sandwich Arches . . . . . . . . . . . . . . . . . . . . . . 127
7.3.2 Sandwich Arch Bridges . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
7.3.3 Cassette Sandwich Arches . . . . . . . . . . . . . . . . . . . . . . . . . 131
7.3.4 Lattice Effects in Sandwich Arches . . . . . . . . . . . . . . . . . . 133
7.3.5 Shear Wall Sandwich Arches . . . . . . . . . . . . . . . . . . . . . . . 134
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
8 Shells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
8.1 Shell Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
8.1.1 Shell Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
8.1.2 New Possibilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
8.2 Cupolas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
8.2.1 Cupolas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
8.2.2 Element Cupolas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
8.2.3 Cupola Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
8.2.4 Cupola Element Design Example . . . . . . . . . . . . . . . . . . . 145
8.2.5 Half Cupolas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
8.2.6 Torus Part Shells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
xiv Contents

8.3 Modular Shells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148

8.3.1 Modular Shell Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
8.3.2 A Modular Shell System . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
8.4 Hyperbolic Shells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
8.4.1 Hyperbolic Shells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
8.4.2 Hyperbolic Element Shells . . . . . . . . . . . . . . . . . . . . . . . . . 151
8.5 Tubes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
8.5.1 Tubular Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
8.5.2 Tunnels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
8.5.3 Drilled Tunnels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
8.5.4 Immersed Tunnels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
8.5.5 Floating Tunnels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
8.6 Floating Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
8.6.1 Floating Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
8.6.2 Sub-Water Cast Floating Structures . . . . . . . . . . . . . . . . . . 157
8.6.3 Concrete Ships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
9 Structural Detailing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
9.1 Anchorage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
9.1.1 Anchorage Basics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
9.1.2 Splitting Strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
9.1.3 Bond Strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
9.1.4 Anchorage Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
9.2 Separable Joints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
9.2.1 Lime Mortar Joints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
9.3 Hinges and Weak Zones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
9.3.1 Concrete Hinges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
9.3.2 Tests of Concrete Hinges . . . . . . . . . . . . . . . . . . . . . . . . . . 169
9.3.3 Weak Zones Guiding Forces and Locating Fracture . . . . 171
9.3.4 Super-Light Concrete Hinges . . . . . . . . . . . . . . . . . . . . . . . 173
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
10 Sustainability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
10.1 Environmental Impact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
10.1.1 Systematic CO2 Assessment . . . . . . . . . . . . . . . . . . . . . . . . 178
10.2 CO2 from Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
10.2.1 Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
10.2.2 Light Aggregate Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . 184
10.2.3 Pumice Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
10.2.4 Foam Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
10.2.5 Aerated Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
10.2.6 Steel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
10.2.7 Recycled Steel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
10.2.8 Timber . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
10.2.9 Other Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
Contents xv

10.3 CO2 from Processes and Transport . . . . . . . . . . . . . . . . . . . . . . . . . . 193

10.3.1 CO2 from Heating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
10.3.2 CO2 from Transport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
10.4 CO2 from Building Components . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
10.4.1 Building Components with 55 dB Sound Insulation . . . . 194
10.4.2 Building Decks with 55 dB Sound Insulation . . . . . . . . . 195
10.4.3 Sand as Sound Insulator for Decks . . . . . . . . . . . . . . . . . . 196
10.4.4 Building Walls with 55 dB Sound Insulation . . . . . . . . . . 196
10.4.5 Building Components with Other Requirements . . . . . . . 198
10.4.6 Sustainability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
11 CO2 Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
About the Authors

Kristian Dahl Hertz M.Sc., Ph.D. is a professor in

design of building structures at the Department of Civil
Engineering at the Technical University of Denmark
(DTU).
After his Ph.D. on Properties of Fire Exposed
Concrete from 1980, he was a part-time associate
professor at DTU and a part-time consulting engineer
for M. Folmer Andersen Ltd. Here, he designed building
structures for the Ministry of Foreign Affairs of Saudi
Arabia, the Royal Theatre in Copenhagen, the National
Bank of Iraq, the Parliament of Greenland, factories and
domestic buildings.
Kristian began teaching in Structural Fire Safety
Design and new courses in structural building design at
DTU. He made research on these subjects in cooperation
with consulting engineers and producers of light aggre-
gate concrete elements. For 16 years, he was a regular
visiting professor at London City University. During
the 1990s, he also served as a head of department and
merged 14 small DTU departments into a Department
of Civil Engineering. At the same time, he was active
in code writing and implemented some of his design
methods in Danish and European codes.
In 2009, Kristian invented super-light structures and
pearl-chain technology and established a course and
a special research area on sustainable light concrete
structures at DTU.
In 2010, three young business persons asked him to
make a spinout company Abeo Ltd based on the ideas,
and the same year, the company won a world champi-
onship in clean technology in San Francisco (Clean Tech
Open).
xvii
xviii About the Authors

A cooperation between DTU, the company, test insti-

tutions, producers, architects, consulting engineers and
contractors developed the new technology and many
practical solutions and details.
They applied the findings in a large number of build-
ings and structures. Factories now produce super-light
concrete elements in the USA and more European
countries.
In 2019, Kristian collected most of the findings about
fire safety of concrete structures in the book “Design
of Fire-resistant Concrete Structures.” In 2020 and
2021, he and Philip Halding collected the findings
on light concrete structures, super-light structures, and
pearl-chain technology in the present book.

Philip Halding M.Sc., Ph.D. is an assistant professor

at the Technical University of Denmark (DTU) in the
Department of Civil Engineering. He is specialized in
sustainable super-light structures and full-scale testing
and monitoring of concrete bridges.
In his Ph.D., he worked with the pearl-chain method
used for bridges, and he was part of the development
from a simple theoretical concept until the erection of the
first-ever pearl-chain bridge. He cooperated with several
industrial partners and another Ph.D. with focus on the
types of concrete used.
After his Ph.D., Philip has worked with advanced
monitoring methods of existing concrete bridges during
full-scale load tests. The purpose of such tests was to
upgrade existing bridges for heavier traffic by wish from
the Danish Road Directorate. The monitoring methods
included non-intrusive wide-angle digital image corre-
lation (DIC), laser scanning and interferometric radar,
as well as a number of other types of equipment.
Philip is the primary teacher of the two DTU master
courses: “super-light structures” and “structural anal-
ysis of buildings.” The course in super-light structures
involves about 80 students in group work with the
different concepts and ideas presented in this book.
Furthermore, Philip is supervising numerous bach-
elor and master student project every semester with
focus on super-light technology. Some of the projects
are in collaboration with the Abeo company, so that the
students work in close collaboration with the industry.
Most of the projects concern actual real-life challenges,
About the Authors xix

where a builder wishes to create a more aesthetical or

sustainable solution, or where the project problem is part
of the current research within super-light structures.
In addition, Philip has graduated the “Pasteur
program,” which is a project management course from
Harvard Business School, and he works actively with
state of the art in teaching methods at university level.
Philip has over the years worked together with Kris-
tian, first with Kristian as a Ph.D. supervisor and later as
colleagues. A large part of the present book is based on
this fruitful research collaboration in the past decade.
Notations

A Area
A Area of openings of a compartment
A1 -A5 Life cycle phases
Ac Area of concrete
Alc Area of light concrete
As Area of steel
Atot Total enclosing surface area of a compartment
Ats Area of crossing steel
B1 -B5 Life cycle phases
C1 -C4 Life cycle phases
D Diameter
D Life cycle phase for reuse
Dc Diameter of strong concrete
Dlc Diameter of light concrete
Dm Outer diameter of concrete with entasis
Ds Diameter of steel
E E-modulus
E0 Initial E-modulus
E0a Initial E-modulus of a concrete number a
E0b Initial E-modulus of a concrete number b
Ea E-modulus of a concrete number a
Eb E-modulus of a concrete number b
Ec E-modulus of concrete
Ec0 Initial E-modulus of concrete
Ec020 Initial E-modulus of concrete at 20 °C
Elc E-modulus of light concrete
Elc0 Initial E-modulus of light concrete
Es E-modulus of steel
Es0 Initial E-modulus of steel
Es020 Initial E-modulus of steel at 20 °C
EA Axial stiffness
EA0 Initial axial stiffness
xxi
xxii Notations

EI Flexural stiffness
EI0 Initial flexural stiffness
EIc0 Initial flexural stiffness of concrete
EIlc0 Initial flexural stiffness of light concrete
EIs0 Initial flexural stiffness of steel
F Force
F0 Strength of centrally loaded elastic column with entasis
FE Euler force
FR Rankine force
FR0 Rankine force of column with entasis
FcE Euler force of concrete in a column
FcR Rankine force of concrete in a column
Fcc Strength of concrete in a cross section
Fcu Ultimate force of concrete in a cross section
Fibot Ring force at bottom of level i of a cupola
Fitop Ring force at top of level i of a cupola
Flc Strength of light concrete in a cross section
FlcE Euler force of light concrete in a column
Flcc Strength of light concrete in a cross section
Flcu Ultimate force of light concrete in a cross section
Fres Resulting ring force at a level of a cupola
Fs Force in steel
Fs1 Force in steel layer 1
Fs2 Force in steel layer 2
FsE Euler force of steel in a column
Fsu Ultimate force of steel in a cross section
Fu Ultimate force (ultimate resistance of a cross section)
H Height of an arch
Hi Height at level i of a cupola
I Moment of inertia
Ic Moment of inertia of concrete
Ilc Moment of inertia of light concrete
Is Moment of inertia of steel
K Parameter
L Length
M Moment
MEnd Moment at end of a beam
Mc Moment in concrete
Mi Moment on an element at level i of a cupola
Ms Moment in steel
N Normal force
O Opening factor
P Single load
Pi Load at level i
R Radius
Notations xxiii

T Temperature
T1 -T64 Temperature parameters
Vi Horizontal reaction at level i
a Parameter
b Parameter
b Parameter for column calculation
b1 Throat height of concrete hinge
b2 Height of concrete hinge
cc Width of concrete cross section
clc Width of light concrete cross section
cp Heat capacity
cs Cover to steel axis
d Cover thickness
dF Depth of force F
di Horizontal distance to load from inclined cupola element
ds1 Depth of steel layer 1
ds2 Depth of steel layer 2
e Eccentricity
fcc Compressive strength of concrete
fcc20 Compressive strength of concrete at 20 °C
fct Tensile strength of concrete
fcu Ultimate strength of concrete
flcu Ultimate strength of light concrete
fs Yield strength of steel
fs20 Yield strength of steel at 20 °C
fsu Ultimate strength of steel
fu Ultimate strength
h Height of cross section
h Height of opening
h Height to a point on a compression arch
hc Height of concrete cross section
hcupola Height of center of cupola
hlc Height of light concrete cross section
k Factor for bond shear strength as part of compressive strength
k Remaining part of strength after cooling to high temperatures
m Parameter
p Load per unit length
q Fire load per unit enclosing surface area of compartment
ri Horizontal radius at level i of a cupola
t Time in minutes
tc Thickness of concrete flange
v Angle parameter
w Width
witop Width at top of cupola element at level i
x Length parameter
xxiv Notations

y Length parameter
yca Height to catenary arch
ycc Length to concrete edge in compression
yci Height to circle arch
yct Length to concrete edge in tension
ypa Height to parabola arch
ys Length to steel axis
α Thermal elongation coefficient
α Angle
δ Deflection
δu Ultimate deflection
ε Strain
εF Strain at a force F
εc20 Strain of concrete at 20 °C
εcu Ultimate strain of concrete
εs Strain of steel
εs20 Strain of steel at 20 °C
εsu Ultimate strain of steel
εsu20 Ultimate strain of steel at 20 °C
εsy Yield strain of steel
εu Ultimate strain
η Stress distribution factor
κ Curvature
λ Conductivity
ν Poisson’s ratio
ξ Reduction of strength
ξc Reduction of concrete strength
ξccCOLD Reduction of compressive strength of concrete in cold condition
ξccHOT Reduction of compressive strength of concrete in hot condition
ξcM Reduction of concrete strength at midpoint
ξs Reduction of steel strength
ρc Density of concrete
σ Stress
σI Biaxial stress in direction I
σII Biaxial stress in direction II
σb Bearing stress
σc1 Concrete stress at side 1
σc1m Concrete stress at side 1 at middle of column
σc1top Concrete stress at side 1 at top of column
σc2 Concrete stress at side 2
σc20 Concrete stress at 20 °C
σc2m Concrete stress at side 2 at middle of column
σc2top Concrete stress at side 2 at top of column
σcau Biaxial concrete strength with hindered lateral expansion
σitop Horizontal stress at top of level i of a cupola
Notations xxv

σs20 Steel stress at 20 °C

σts Steel stress 31.5 MPa at ultimate tensile strain of concrete
τ Shear stress
Chapter 1
History

Abstract The historical development of concrete and light concrete is explained

and dates all the way back to the Phoenicians at approximately 1200 BC. Super-light
concrete technology is also inspired by techniques of vaults and cupolas developed
in the Roman Empire.

1.1 Ancient Building Materials

1.1.1 Gypsum and Limestone Mortar

Now and then, you hear a theory that the pyramids of Giza in Egypt (2580 BC)
were made of concrete. That is not true. The largest of them are made of blocks of
limestone placed in a mortar that according to the author’s investigations primarily
is made of gypsum with small grains of burned clay. An orange-red colour of the
grains caused by the content of iron oxide shows that the material has been burned
at 400 °C [1, 2], where less than 300 °C is needed for burning a gypsum mortar [3].
This means that the Egyptians, for whom wood was a limited resource, could
produce gypsum mortar with less energy than needed for burning limestone (900 °C)
and for making a cement. Furthermore, builders were familiar with gypsum at that
time, since you can trace it as a building material to 4.000 BC in the Sumerian city
Uruk [4] and perhaps even back to 9.000 BC in Catalhöyük in Turkey [5].
However, the pyramids can tell us something else of importance for the later
concrete production. If you measure their orientation, you will find that all 80 pyra-
mids in the Northern Egypt are rotated about four degrees with respect to North
(Fig. 1.1).
North is in the middle between the direction where the sun rises and where it sets,
so the Egyptians had no problems in determining that. (Explanation follows).
When the desert developed in the northern Africa and Middle East, people had to
concentrate at the rivers. The first strong societies therefore appeared here, and the
rulers needed a way to demonstrate their power.
King Djoser’s architect Imhoteph placed the first step pyramid in Saccara at the
Nile about 2660 BC (Fig. 1.2). He probably intended to make a copy of mountains

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 1

K. D. Hertz and P. Halding., Sustainable Light Concrete Structures, Springer Tracts
in Civil Engineering, https://doi.org/10.1007/978-3-030-80500-5_1
2 1 History

Fig. 1.1 Pyramids in Giza. Photo KD Hertz

Fig. 1.2 King Djoser’s step pyramid. Photo KD Hertz

with the same shape, which serve as landmarks.

(You may for example find one in the desert East of Oman shown in the film
“Lawrence of Arabia”, taken on location in 1962).
Wind from North and West have eroded soft layers of these natural mountains, so
that they became step pyramids similar to the artificial one he built.
We can also measure this first pyramid to be rotated four degrees with respect to
North.
1.1 Ancient Building Materials 3

The natural preconditions for developing new building materials were better in
other areas of the ancient world.
The Phoenicians lived for example in Lebanon and northern Israel and had rich
amounts of fuel from the forests. They burned limestone and clay for mortar, tiles,
and bricks and they had a large production of glass. They exported glass pearls, vases,
and raw glass bars to the entire Mediterranean.

1.1.2 Ancient Concrete

The Phoenicians made the oldest concrete that the authors know about approximately
1200 BC. As mentioned above, they had a large glass production (Fig. 1.3). They
pulverized glass and mixed it with burned limestone and water creating calcium-
silicate-hydrate. A material, which we know as hydrated cement. They also made
cement from pulverized tiles and burned limestone.
The Bible tells us that the Phoenician king gave King David some masons, when he
founded Jerusalem. That was not just a helping hand. It was technology transfer intro-
ducing concrete as a new building material. Today, you can see concrete claddings
of water cisterns made from this material in Jerusalem [5].
Later, the Phoenicians learned to use fly ash from the volcano at Santorini replacing
the more expensive pulverized glass and tile in their concrete. At this island, you can
see Phoenician fly ash quarries next to Roman quarries showing us how the Romans
may have learned about concrete from the Phoenicians.
The Romans identified fly ash applicable for concrete production from volcanoes
at many locations in their empire. Vitruvius talks about fire in the earth [6].

Fig. 1.3 Phoenician glass. Photo KD Hertz

4 1 History

This has something to do with the orientation of the pyramids.

The reason why we have these volcanoes on a line in the direction East–West
through the Mediterranean is that the whole continent of Africa has turned four
degrees since they were built. With Gibraltar as centre of rotation, the African tectonic
plate dives under the European creating a basis for volcanic activity. It is amazing
that we can read the movement of continents on man-made structures.
In Italy, the Romans found plenty of fly ash especially from the European super-
volcano Campi Flegrei (Fig. 1.4) at Pozzuoli near Naples. This city has given the name
Pozzolana to the siliceous materials reacting with calcium oxide (burned limestone)
and water to calcium silicate hydrate (hydrated cement). Big eruptions of this super
volcano at 35.000 BC and 10.000 BC gave volcanic ash especially to the East as far
as in southern Russia.
This and materials from other volcanoes became a basis for the Roman concrete
production.
In southern Germany, they found tuff that was applicable. In areas with none
or sparse occurrence of natural Pozzolana, they applied the old recipes based on
pulverized tiles.
They could also mix this with natural fly ash. Furthermore, they mixed cement
with aggregates of stones, pieces of rock, or crushed parts of old concrete or tile
structures to make the new concrete also known as opus caementicium.

Fig. 1.4 Campi Flegrei. From the crater of the European super volcano. Photo KD Hertz
1.2 Ancient Structures 5

1.2 Ancient Structures

1.2.1 Walls

For walls and columns, the Romans cast concrete between surface shells of bricks
or stones, which they masoned simultaneously with casting the concrete.
Besides from participating in the load-bearing function of this composite structure,
the shells served as permanent moulds and as endurable surfaces protecting the
concrete against mechanical impact, weather, and heat from fire.
Roman builders applied mainly three different kinds of permanent moulds.
Opus Incertum consists of irregular stones placed as the plane surfaces keeping
the concrete core in place. The Roman wall next to Tower of London is an example
of that.
For nice surfaces, Roman builders applied Opus Latericium (Fig. 1.5) that is
masonry of flat bricks of burned clay.
They often gave the flat bricks a pentagonal shape with a rectangular part at the
surface and a triangular part pointing into the central concrete material to obtain a
better grip. You typically find Opus Latericium in facades and on ends of internal
walls, where they become visible.
They used Opus Reticulatum as shown in Fig. 1.6 mainly for internal walls.
It was fast to build, because the builder could easily place the long pyramidal
ceramic tiles with quadratic bottom in the concrete and simultaneously adjust
skewness and unevenness.

Fig. 1.5 Opus Latericium. Photo KD Hertz

6 1 History

Fig. 1.6 Opus Reticulatum. Photo KD Hertz

When Phoenicians and Romans applied fly ash from the volcanoes in their
concrete, it is only natural that they also applied volcanic stones including pumice
as aggregate.
This means that they applied what we today call light aggregate concrete and
they often used several concrete qualities in the same structure to obtain a suitable
distribution of weight and to guide the forces in the structure to where they wanted
them to be and to avoid tension as far as possible [7, 8].
Application of light aggregate concrete reduces the dead load and thereby the
pressure on the permanent moulds. As mentioned, they could also apply pieces of
tile as aggregate, and sometimes they used hollow pots or vases as aggregate.
By means of this, they were able to include larger air filled volumes in order
to obtain a lower weight and density of the concrete structure than possible by
application of natural pumice as light aggregate.
The pots of burned clay also had a certain strength, which means that the master
builder obtained a weight reduction without suffering the same strength reduction as
would follow if he applied a natural pumice aggregate to get the same low density.

1.2.2 Arches and Vaults

Builders have applied arches since pre historic time. Arches are capable of with-
standing uniformly distributed load in a perfect way (without bending moments). The
main benefit is that you may obtain a certain span only by application of compression
forces, which you can guide from load to support by means of unreinforced structures
of stone and concrete.
1.2 Ancient Structures 7

In the Egyptian museum in Cairo, you can see model toy houses of burned tile
from 2500 BC, where the buildings are made of vaults, which give lateral support to
each other (Fig. 1.7).
You could imagine that the ancient builders made their multi vault structures like
present day builders do it in Sudan. Figure 1.8 shows an example of the technique.
First, they make the walls of the houses and their gables. Then, they place mud
bricks, fibre reinforced with straw, as inclined arches leaning against the gable and

Fig. 1.7 Egyptian toy house 2500 BC. Photo KD Hertz

Fig. 1.8 Vaults in Sudan made without scaffolding. Photo KD Hertz

8 1 History

Fig. 1.9 Granaries in Ramesseum in Luxor 1400 BC. Photo KD Hertz

therefore without any scaffolding! Finally, they finish the vault with an outer mortar,
or they may apply a second layer of brick arches leaning against the opposite gable.
The builders applied the same basic vault building principles in the granaries of
the Ramesseum in Luxor from about 1400 BC (Fig. 1.9). Here the approximately
100 m long vaults also support each other laterally. Each vault span about 8 m and
consists of four layers of mud bricks, which are fibre reinforced with straw.
Roman builders developed arch building technology and applied it for many
purposes.
In a bridge like Ponte Fabricio (Fig. 1.10) the master builder made two free spans
of 24.5 m to reach the Tiber island—Isola san Bartolomeo—in Rome.
The bridge is from year 62 BC, and history tells us that the master builder had to
wait 20 years for his payment, because the authorities would like to be sure that it
could stand the impact of the river.
Roman builders applied concrete vaults covering paths and staircases in most
public buildings.
They also applied vaults of considerable dimensions for side bay of basilicas.
For example, a full symphony orchestra can find place in one of the six side bays
of the Basilica of Maxentius and Constantin in Rome.
One important application of the ancient vault is as fire protection.
A vault carries its load by leading compressive forces through fire resistant mate-
rials as concrete and brick. A possible shrinkage or deformation of these materials
may cause a minor settlement but not a major collapse.
Roman builders therefore applied them for ceilings over shops at the bottom of
their buildings separating the combustible content of the shops from the upper stories,
where other people lived, and where they might apply lighter and more flammable
wooden floors.
1.2 Ancient Structures 9

Fig. 1.10 Ponte Fabricio. Photo KD Hertz

Isolated staircases of stone, brick, or concrete lead to the upper stories.

The builders also usually installed a wooden floor under the vault of the shop, so
that the owner’s family could sleep there and serve as a living fire alarm.
Figure 1.11 shows an elegant Roman concrete vault covering a circular path. Here,
the master builder has created a vault with an extremely small thickness of only few
centimetre, and constructed a horizontal ring beam where the lateral forces from the

Fig. 1.11 Ultra-thin concrete vault Villa Adriana. Photo KD Hertz

10 1 History

Fig. 1.12 Horizontal vault Trajan’s Market. Photo KD Hertz

vault is counteracted by compression in the ring so that the columns should only
carry vertical reactions.
Figure 1.12 shows another elegant application of a vault. Here the master builder
(Apollodorus from Damascus) shaped the entire eight storey high building of Trajan’s
Market in Rome as a big vault resisting the lateral pressure from the mountain that
was carved out in order to make space for the market.

1.2.3 Cupolas

Application of light aggregate concrete based on pumice reduces the dead load. The
best Roman master builders took advantage of the light aggregate concrete in their
shell structures (such as vaults and cupolas), where the reduction of the dead load
opened possibilities for creating large span widths. As mentioned above, they could
also mix pots of burned clay in the concrete in order to reduce the density by creating
empty volumes.
Figure 1.13 shows a cross-section of a cupola from Hadrian’s villa outside Rome.
As for many other cupolas, the master builder of this one constructed a horizontal
round hole at the top, where the forces become small.
Sometimes, as for example in the Trajan’s Market in Rome, a horizontal compres-
sion ring of bricks provides the round hole with a well-defined edge, and it serves to
counteract a possible residue of radial compression forces.
The relatively thick walls and pillars beneath the cupola lead the horizontal
reactions of the cupola further into the ground.
1.2 Ancient Structures 11

Fig. 1.13 Light concrete shell Villa Adriana. Photo KD Hertz

As also seen in Fig. 1.13, a series of arches support the cupola and lead the forces
down in pillars between openings and niches in the wall below.
You may often find such relieving arches in walls beneath cupolas, even when
you do not find openings or niches. This is for example seen several places in the
walls supporting the cupola of Pantheon in Rome.
We know that the Romans could apply several layers of light concrete of different
density in the same structure. They did for example apply no less than eight horizontal
layers of light concrete with different densities in the construction of the Pantheon.
By means of that, they ensured that the compressive forces were kept within the
cross-sections for the large span of 43 m [9].
Apollodorus of Damascus, who rebuilt the temple and constructed the present
cupola, designed this light concrete shell with cassettes (deep panels with less mate-
rial thickness, also known as coffers) (Fig. 1.14). By doing so, he further reduced
the weight compared to its load bearing capacity and simultaneously, he guided the
forces into the radial and circular ribs.
You find other delicate, load-bearing shell structures in the many half cupolas
of the Roman baths. Today, we can only guess how they were able to design them.
Modern engineers have a severe challenge if they try to construct similar buildings
and estimate the stress distributions without the aid of computer software. (See 8.2.4).

1.2.4 Roman Reinforced Concrete Structures

The Roman builders applied concrete and light concrete in structures, where forces
act in compression. Arches and vaults lead vertical loads to the sides where thick
12 1 History

Fig. 1.14 Pantheon Cupola. Photo KD Hertz

walls or massive sections withstood the lateral forces from the arches. The walls
were hollowed out with niches or supported by cross walls.
Nevertheless, one particular structure called for application of flat concrete slabs
subjected to bending: “Hypocaust”—a type of floor heating system.
They lead hot air under the floors in bath buildings, and from there the system
could guide it up via ceramic pipes embedded in the walls.
For such floors, vaults were not so applicable because the height could be limited
and because a vault would capture the hot air and hinder sideward’s distribution of
it.
Flat concrete slabs were therefore preferable.
This lead to a structural problem, since such flat slabs act in bending for which
plain concrete cannot resist much tension stresses.
The Romans therefore in general had to support the flat concrete slabs with piles
of brick at a small distance from one another as shown in Fig. 1.15.
However, the authors know about at least one structure showing another solution
to this problem.
In a slab covering a floor-heating system in the so-called “Repräsentationshaus”
at Magdalensberg near Klagenfurt in Austria, the Roman builder introduced an iron
reinforcement to take tension forces in the bottom side [10].
Here, the builder placed a series of 1800 mm long flat iron bars with cross-section 5
times 30 mm and a 200 mm spacing to resist the tension forces in the bottom of a
160 mm thick concrete slab covering a 1070 mm wide hot air channel [5], (Fig. 1.16).
The Romans called the Klagenfurt area Noricum. It was famous for steel
production.
1.2 Ancient Structures 13

Fig. 1.15 Roman hypocaust system. Photo KD Hertz

Fig. 1.16 Norican iron reinforcement

Each year merchants from the whole empire came here to buy tools and weapons
of the best quality. It is therefore no wonder that we find the first steel reinforcement
in a concrete structure here.
The Roman builders also applied fibre reinforcement to take tensile stresses or
to distribute cracks from tensile forces. Fibre reinforcement could be many different
qualities, but the builders often applied hairs from horses or donkeys.
Making aqueducts, builders had the task to keep the structure watertight, also
when the terrain was setting or deforming. They could therefore apply up to eight
different layers of mortar (opus signinum) in the channel.
In such hydraulic structures, builders could make the mortar near the water tight
with olive oil. In order to avoid cracks in this dense layer, they could place one or
more layers of mortar with fibre reinforcement that distributed cracks from the deeper
layers.
14 1 History

Fig. 1.17 Aqueduct and pressure pipe at Aspendos in Turkey. Photo KD Hertz

The inmost mortars often contained red pulverized burned clay. The next layers
often contained pulverized marble. Then followed layers with charcoal. This made
a typical sequence of red, white, and grey mortar layers [11].
Sometimes the builders placed the water channel on a layer of small arches, which
served two purposes. To create the inclination wanted and to distribute settings from
the larger arches with longer spans below.
In some places, the builders applied pressure pipes to avoid building a high aque-
duct for example across a wide valley as that in Aspendos. Here, it was a problem,
that the dynamic inertia of the water flow might give a blow destroying the pipe if
something interrupted the flow. Today, we know this effect as a “hydraulic ram”.
At Aspendos, the master builder solved this problem by making the aqueduct in
full height at two points as seen in the Fig. 1.17 and placing an open ventilated basin
at the top of each.
In other locations, they solved the problem by providing pressure pipes with safety
valves. They made these as two or three holes in the strong concrete pipe filled with
a lighter concrete of less strength.

References

1. Hertz KD (1980) Betonkonstruktioners Brandtekniske Egenskaber. (Properties of Concrete

Structures Exposed to Fire). (In Danish). Department of Building Design, Technical University
of Denmark. 209p
2. Hertz KD (2019) Design of fire-resistant concrete structures. ICE-Publishing, Thomas Telford
Ltd. ISBN: 9780727764447. 248p. London 2019
References 15

3. Eckel EC (1905) Cements, Limes and plasters. Their materials, manufacture, and properties.
Wiley, New York. 698p
4. Haegermann G (1964) Vom Cæmentum zum Spannbeton, Part 1. (In German). Bauverlag
GmbH, Wiesbaden-Berlin. 491p
5. Stark J, Wicht B (1998) Geschicte der Baustoffe. (In German). Bauverlag, Wiesbaden und
Berlin 205p
6. Vitruvius (23 BC) The Ten Books on Architecture. By Morgan MH (1960) (In English) Dover
New York, 331p, and by Isager J (2016) Om Arkitektur (In Danish) South Danish University
Publishing, Odense, Danmark. 505p
7. Hertz KD (2010) A new patented building technology based on ancient Roman knowledge. In:
Proceedings of symposium “handling exceptions in structural engineering” La Sapienza, 6 p
Rome, Italy, 8–9 July 2010
8. Hertz KD, Schmidt JW, Goltermann P (2014) Super-light and pearl-chain technology for
support of ancient structures. In: Proceedings of the 2’nd international conference on protection
of historical constructions PROHITECH’14, pp 641–646, May 2014, Antalya, Turkey
9. de Fine Licht KR (1968) The Rotunda in Rome. Jutland Archaeological Society Publications,
A Study on Hadrian’s Pantheon, 344p
10. Lamprecht HO (1985) Opus Caementitium Bautechnik der Römer. (In German) Beton-Verlag
Düsseldorf
11. Frontius SJ (98 AD) De Aquae Ductu Urbis Romae. (About the Water Supply of the City of
Rome) Translated by Hansen J (In Danish) 1982 Museum Tusculanum 215p
Chapter 2
Materials

Abstract Super-light structures utilizes concrete types of different strengths and

densities. Knowing the properties of the different concrete types can enable engineers
to build lighter and with fire safety. The chapter also provides information about the
use of reinforcement steel and pre-stressing.

In this chapter, we present data for a number of materials relevant for the structures
in the book. We do that in order to give the reader an idea of which materials, we
have in mind for the structures presented.
Designers can use the data for preliminary calculations until they choose the actual
materials for their structures with respect to local conditions such as availability of
natural minerals, price levels, production methods, and local codes, legislation, and
traditions.
We recommend you to read about the history and origin of building materials in
Chap. 1 first, because this may give you a deeper understanding of their nature and
detailed composition.
Figures 2.1 and 3.3 show an example of an advanced application of reinforced
concrete for an optimal structure by Piere Luigi Nervi [1, 2].

2.1 Heavy Concrete

2.1.1 Ordinary Concrete

The concept of concrete comprises a variety of materials with different aggregates,

cement qualities, mixtures, strengths, densities and other physical properties.
Aggregates depend usually on available local resources as for example granite,
basalt, limestone, and quartz. Producers can also apply other natural or artificial
aggregates like pumice and ceramics.
Cement can vary for example depending on the local availability of lime.
Producers and contractors can substitute it partly or fully by artificial or natural

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 17

K. D. Hertz and P. Halding., Sustainable Light Concrete Structures, Springer Tracts
in Civil Engineering, https://doi.org/10.1007/978-3-030-80500-5_2
18 2 Materials

Fig. 2.1 Palazzetto dello sport by PL Nervi 1958. Photo KD Hertz

pozzolana like fly ash from volcanoes and chimneys or crushed ceramics, and they
can apply a great variety of chemical additives influencing the physical properties
and workability of the concrete.
For the purpose of this book, we need to identify some groups of concrete materials
and apply them for different purposes. A heavy concrete has usually a density of at
2300 kg/m3 or more. In this book, we call a concrete heavy, if the density is at least
2300 kg/m2 . We call all qualities of less density light. This definition accord with
the distinction applied in most other codes and textbooks.
Some concrete codes confine themselves to certain intervals of strength. The
Eurocode [3] for example apply an upper limit of strength of 60 MPa.
In practise, engineers often apply stronger materials, but treat them according
to the codes by using a calculational strength within the limit instead of the real
measured physical strength.
In this book, we call a heavy concrete, which is defined to be within 60 MPa, for
ordinary concrete or just concrete in contradiction to a high-strength concrete (See
Sect. 2.1.2).
In practise, you should find and document the material properties by tests.
However, as a guideline for preliminary calculations and student projects, we give
some typical values in Table 2.1. They are all based on 5% quantiles (95% of a series

Table 2.1 Concrete data

Compressive strength fcc 10 20 30 40 50 60 MPa
Tensile strength fct 1.0 1.5 2.0 2.5 2.9 3.1 MPa
Initial E-modulus Ec0 22 31 36 38 40 42 GPa
2.1 Heavy Concrete 19

of tests has higher strength or stiffness) from own not reported test results, supported
by CEN [3], DS [4].
For many applications of heavy concrete in super-light elements dealt with in
this book, we apply a concrete with fcc = 55 MPa, fct = 3.0 MPa, Ec0 = 41 GPa,
thermal elongation coefficient α = 1.1·10–5 /°C, heat capacity cp = 1.00 kJ/m3 , and
conductivity λ = 0.90 W/m°C.
Most textbooks and codes on concrete structures apply idealized stress–strain
curves for concrete. They often give formulas for the curves, which look quite
different, because they are semi-empirical and apply additional old-fashioned
assumptions such as a ratio 1000 between the initial E-modulus and the strength.
However, you can often identify the curves as derived from the same basic assump-
tion made by Ritter. According to him, you can find the E-modulus of a concrete at
a certain compressive stress as the initial E-modulus reduced by the ratio between
the stress and the strength [5].
This means, that the E-modulus Ec at a certain compressive stress σ on a concrete
with a compressive strength fcc at an ultimate strain εcu (which is usually 0.35%) can
be expressed as

Ec = Ec0 (1 − σ/fcc ) or dσ/dε = Ec0 (1 − σ/fcc )

Solving this differential equation, you find the following expression for the
working curve

σ(ε) = fcc (1−e−(Ec0 /fcc )ε )

Figure 2.2 shows the curve.

If the concrete is heated up for example by a fire exposure, aggregates expand
and cement paste releases physically and chemically bound water and shrinks. The
concrete will therefore become more porous, crack, and lose strength, stiffness and
conductivity. These reductions will therefore mainly depend on the aggregate applied
[6].

Fig. 2.2 Idealized

stress–strain curve for
concrete
20 2 Materials

When the concrete cools down at the end of a fire, the cement paste will absorb
water again, and new calcium hydrate crystals will widen up the cracks.
As a result, strength and stiffness will reduce further in the cooling phase. We
therefore distinguish between the reduction ξccHOT (T) of the concrete compressive
strength at a certain temperature T in a HOT condition during a fire and the final
reduction ξccCOLD (T) in a COLD condition after the concrete has cooled down from
a maximum temperature T.
See detailed explanations and methods for fire-safety design of concrete structures
in the textbook “Design of Fire-resistant Concrete Structures” [7].
Strength reduction as a function of temperature (in °C) will be an S-shaped curve,
and for all concrete and steel materials. You can calculate it by the same general
formula
1−k
ξ(T) = k +    2  8  64
1+ T
T1
+ TT2 + TT8 + TT64

k is the part of a steel strength regained after cooling, and it is always 0 for
concrete.
T1 . T2 , T8 , and T64 are constants in degree Celcius for the actual material. You
find the values in Table 2.2. The E-modulus of a concrete will be reduced by the
square of the strength reduction. This means that

fcc (T) = ξc (T)fcc20 , Ec0 (T) = ξc (T)2 Ec020 and εcu (T) = 0.35%/ξc (T)

Any point (εc20 ,σc20 ) at the stress–strain curve is transferred to

(εc20 /ξc (T), ξc (T) · σc20 ).
The temperature in a HOT condition and the maximum temperature during the
fire, which you apply in a COLD condition, vary through a concrete cross-section.
You can consider this variation by using a reduced cross-section, where you apply the
actual strength reduction ξcM in the midpoint. You reduce the width by subtraction
of damaged zones at the fire-exposed surfaces. The reduction factor η for the width
represents the average reduction of strength related to that in the midpoint [7]. We
call the value η the stress-distribution factor. It is between 0 and 1.0.
You can find free programs for the strength reduction (“Damage”) and for the
temperature distribution and average strength reduction (“ConFire”) at the homepage
of the Department of Civil Engineering [8], where you also find User’s guides for
the programs.
The program ConFire calculates temperature and strength reduction of a rein-
forcing bar or a concrete in a specific point of a cross-section at a specific time t
in minutes. This is based on the time of max temperature during a fire (the HOT
condition) and the max temperatures during a full fire course (the COLD condition).
The fire courses can be a standard fire according to [9] or a fully developed fire.
The latter is defined by a fire load q in MJ per m2 enclosing surface Atot , and an
opening factor O in m½ depending on the areas of openings A and their average
2.1 Heavy Concrete 21

Table 2.2 Material data for fire-reduction

k T1 T2 T8 T64
Siliceous concreteHOT 0.00 15,000 800 570 100,000
Siliceous concrete COLD 0.00 3500 600 480 680
Main group concreteHOT 0.00 100,000 1080 690 1000
Main group concrete COLD 0.00 10,000 780 490 100,000
Light aggregate concreteHOT 0.00 100,000 1100 800 940
Light aggregate concrete COLD 0.00 4000 650 830 930
Russian Fire resistant concrete on Chamotte 0.00 100,000 100,000 1150 1150
HOT + COLD
Danish Fire resistant concrete on Mo-Clay 0.00 10,000 4000 710 1100
HOT + COLD
Hot rolled bars 0.2% stressHOT 0.000 6000 620 565 1100
Hot rolled bars 2.0% stressHOT 0.000 100,000 100,000 593 100,000
Hot rolled bars 0.2% stress COLD 1.000 100,000 100,000 100,000 100,000
Hot rolled bars 2.0% stress COLD 1.000 100,000 100,000 100,000 100,000
Cold worked bars 0.2% stressHOT 0.000 100,000 900 555 100,000
Cold worked bars 2.0% stressHOT 0.000 100,000 5000 560 100,000
Cold worked bars 0.2% stress COLD 0.580 100,000 5000 590 730
Cold worked bars 2.0% stress COLD 0.520 100,000 1500 580 650
C-w prestressing steel 0.2% stressHOT 0.000 2000 360 430 100,000
C-w prestressing steel 2.0% stressHOT 0.000 100,000 490 450 100,000
C-w prestressing steel 0.2% stress COLD 0.200 100,000 750 550 650
Quenched and Tempered 1500 MPa 0.2% 0.000 1100 100,000 430 100,000
stressHOT
Quenched and Tempered 1500 MPa 2.0% 0.000 3000 1400 450 100,000
stressHOT
Quenched and Tempered 1500 MPa 0.2% 0.213 100,000 10,000 590 660
stress COLD
Quenched and Tempered 1500 MPa 2.0% 0.213 100,000 10,000 590 660
stress COLD
Quenched and Self-tempered 550 MPa 0.2% 0.000 6000 1150 540 700
stressHOT
Quenched and Self-tempered 550 MPa 2.0% 0.000 100,000 100,000 590 700
stressHOT
Quenched and Self-tempered 550 MPa 0.2% 0.418 100,000 100,000 700 900
stress COLD
Quenched and Self-tempered 550 MPa 2.0% 0.437 100,000 100,000 700 900
stress COLD
22 2 Materials

opening heights h as
√
A h
O=
Atot

The cross-section can be insulated for example by an insulation material or by a

light concrete. This means that you can also apply the program for fire safety design
of a composite structure or a super-light structure as presented in Chap. 3.
For CO2 emissions, please consult Chap. 11 on Sustainability.

2.1.2 High-Strength Concrete

Some concrete codes like the Eurocode [3, 10] define as mentioned in Sect. 2.1.1, a
strength of approximately 60 MPa as an upper limit for their application.
Researchers have for many years tried to develop concrete materials of a substan-
tially higher strength by filling out the space between the cement grains with fine
particles. The main hindrance they had for achieving that was electrical charges
hindering them to place the particles that close.
However, about 1980 Hans Henrik Bache succeeded in doing so using the newly
developed super-plasticizing additives [11]. These additives neutralize the electrical
charges and make the concrete easier to cast and, as Bache realized, opened a possi-
bility for making a more compact material. By means of this new technology, he
could fill out the cavities between the cement grains with silica fume or other small
particles.
He was able to give his new concrete a compressive strength of 240 MPa and he
called it “Densit”. He applied burned bauxite as an aggregate since the strength now
became so high that the strength of ordinary granite would be a limiting problem.
The colour was black, because carbon adhered to the fly ash.
The first author Hertz found the new strong materials promising for design of
light structures with long spans and small cross-sections. He therefore asked Bache
to send a pallet of test cylinders, so that he could investigate the properties of the
new material and find if there were any limitations for the application of it.
The strong and stiff material gave rise to violent ruptures splitting the cylindrical
specimens vertically when testing their compressive strengths. Hertz had foreseen
that and provided the test machine with a safety screen. However, what he had not
foreseen was that the material exploded when being slowly heated up by only one
degree Celsius per minute in an oven in order to determine the residual compressive
strength after heating and cooling.
The slow heating rate was chosen in order to avoid destructive thermal stresses in
the specimens, but the test cylinders exploded at only 300 ºC. Figure 2.3 shows an
explosive fraction of one of the first cylinders tested.
It was immediately obvious, that the steam pressure trapped in the extremely
dense material caused the explosion. [12–17].
2.1 Heavy Concrete 23

Fig. 2.3 Exploded cylinder of Densit. Photo KD Hertz

The author also dried test specimens out at 100 ºC and checked by weight that
they did not contain any free water. These specimens did also explode proving that
crystal water is sufficient for explosive spalling of such a dense concrete.
Explosive spalling appears to be a paramount hindrance for application of high
strength concrete in buildings, where fire exposure is relevant. It has therefore been
a subject for further investigations leading to the following limits for application of
concrete in building design as given by [7]:
• Dry light-aggregate concrete with open pores is not susceptible to explosive
spalling.
• Ordinary concrete of strength less than 80 MPa may suffer from spalling of corners
or sides of fresh or wet structures within the first 20 min of a standard fire.
• High strength concrete stronger than 80 MPa subjected to compressive stresses
or hindered thermal expansion is susceptible for explosive spalling.
• Free unloaded structures of ultra-high strength concrete of more than 150 MPa
may spall explosively.
This leads to the following more general rule for application of concrete, where
fire-exposure is relevant

A concrete appears to be safe concerning explosive spalling, if its compres-

sive strength is less than 60 MPa and its moisture content is less than 3%
by weight.
If a higher strength is applied, the temperature should not exceed 350 ºC.
24 2 Materials

Table 2.3 High-strength concrete data

Compressive strength fcc 100 150 240 300 400 MPa
Tensile strength fct 4.0 8.0 10 11 12 MPa
Initial E-modulus Ec0 43 45 50 55 60 GPa
Density ρc 2350 2400 2500 2550 2700 kg/m3

By embedding the high-strength concrete into a light and insulating concrete, you
may fulfil the conditions above.
Super-light and composite structures dealt with in this book therefore represent
a possibility for utilization of the benefits of high-strength concrete that elsewise
cannot be applied.
This can lead to lighter structures and longer spans etc. than whatcan be obtained
with ordinary concrete structures.
High-strength concrete is developed towards still higher strengths. Since Bache’s
Densite of 240 MPa, strengths of 300−, 400−, and even 800 MPa have been achieved.
Table 2.3 shows examples of properties of high-strength concrete, which are
known to be applied. They are based on 5% quantiles. Considerable deviations
may occur depending on the actual recipes, aggregates, choice of small particles,
additives, and the procedure applied when producing the concrete.
For some applications of high-strength concrete in super-light elements dealt with
in this book, we apply a material with fcc = 110 MPa, fct = 5.0 MPa, Ec0 = 43 GPa,
density 2360 kg/m3 , thermal elongation coefficient α = 1.1·10–5 /°C, heat capacity
cp = 1.00 kJ/m3 , and conductivity λ = 0.90 W/m°C.
For other applications, where the denseness is of importance as for example a
strong outer surface layer of a tunnel, a floating building, or ships dealt with in
Sects. 9.5 and 9.6, we consider high-strength concretes at the level of the original
Densit.
It could be one with fcc = 300 MPa, fct = 11 MPa, Ec0 = 55 GPa, density
2550 kg/m3 , thermal elongation coefficient α = 1.1·10–5 /°C, heat capacity cp =
1.00 kJ/m3 , and conductivity λ = 1.00 W/m°C.

2.2 Light Concrete

2.2.1 Light Aggregate Concrete

As mentioned in Sect. 2.1.1, you can reduce the density of concrete by application
of light aggregates.
Natural pumice from volcanoes is still widely applied, as it was in antiquity (see
Sect. 1.2.3).
It may be transported long distances by ship. In Denmark, light aggregate concrete
producers apply pumice from Iceland.
2.2 Light Concrete 25

Other alternatives are expanded clay or expanded perlite. The quality of these
aggregates varies with the raw materials and the processes applied by the producer.
Figure 2.4 shows a cylinder of a light aggregate concrete of density 700 kg/m3
commonly applied for super-light structures (see Chap. 3).
The aggregate is expanded clay and it has a compressive strength of 3.5 MPa
and a tensile strength of 0.30 MPa, initial E-modulus 3.5 GPa, thermal elongation
coefficient α = 1.1·10–5 /°C, heat capacity cp = 1.00 kJ/m3 , and conductivity λ =
0.35 W/m°C.
As you can see, the concrete is quite porous, which means that there is no risk of
explosive spalling, and that you can apply the material for sound damping surfaces
as explained in Chap. 4.
Figure 2.5 shows ovens applied for production of expanded clay aggregates by
a Danish factory. As explained in 10.2.2, the factory has now replaced 80–90% of
the coal applied for heating the ovens with CO2 neutral waste. The CO2 impact is
therefore halved from 0.27 to 0.14 kg CO2 per kg of the resulting light-aggregate
concrete.
Table 2.4 shows some possible material properties based on 5% quantiles.
Light-aggregate concretes represent considerable deviations mainly due to different
qualities of the aggregates (Fig. 2.4).

Fig. 2.4 Light aggregate concrete. Photo KD Hertz

26 2 Materials

Fig. 2.5 Expanding clay aggregate ovens. Photo KD Hertz

Table 2.4 Light-aggregate concrete data

Density ρc 600 700 900 1200 1800 kg/m3
Compressive strength fcc 3.00 3.5 5.0 7.5 15.0 MPa
Tensile strength fct 0.20 0.30 0.8 1.8 2.5 MPa
Initial E-modulus Ec0 3.00 3.5 4.5 7.0 16.5 GPa
Conductivity λ 0.30 0.35 0.40 0.45 0.90 W/m°C

2.2.2 Foam Concrete

Foam concrete has proven to be an applicable material in remote areas, where it may
be difficult to get light aggregates. However, the recent development unveils several
promising possibilities for the material.
You simply create a foam and mix it gently into a cement slurry. Even barber foam
can be applied, but more robust foam liquids are developed for the purpose. Some
companies like Aercrete produce transportable foam concrete mixing machines.
Table 2.5 shows some examples of foam concrete data from [18, 19].
Foam concrete has a fire-resistance equal to the one of light-aggregate concrete
(Fig. 2.6). You can therefore estimate fire safety for structures with foam concrete
using values for light-aggregate concrete in Table 2.2.
In practise, densities of above about 600 kg/m3 are applied for load-bearing struc-
tures. However, at the author’s university a number of projects have developed much
lighter foam concretes. Such foam concrete is not for structural applications, but
created as an insulating material, which can be cast in place and fill out all kinds of
cavities.
2.2 Light Concrete 27

Table 2.5 Light foam concrete data

Density ρc 100 300 600 700 900 1200 1800 kg/m3
Compressive strength fcc - 1.00 3.00 3.5 5.0 7.5 15.0 MPa
Tensile strength fct – 0.05 0.20 0.30 0.8 1.8 2.5 MPa
Initial E-modulus Ec0 – 1.00 3.00 3.5 7.0 7.0 16.5 GPa
Conductivity λ 0.035 0.05 0.11 0.12 0.20 0.32 0.85 W/m°C

Fig. 2.6 Foam concrete. Photo KD Hertz

Jensen and Vahlgren [19] saw this potential and developed insulating foam
concretes down to 200 kg/m3 . Later, others have followed, at qualities with a density
of 100 kg/m3 and conductivities comparable with that of mineral wool insulation
have been obtained.

2.2.3 Aerated Concrete

Aerated concrete is widely applied for industrially produced building blocks and
wall elements. The factory adds an aluminium powder to cement slurry and fill it
into a train of small wagons on a track.
The slurry expands due to formation of bubbles and after some hardening, the
volumes at the wagons are cut into blocks of the size wanted.
The wagons are then driven into an autoclave, which can be a tube of length 100 m
in which the material is subjected to heat and pressure for some time. This process
stabilizes the aerated concrete.
28 2 Materials

A typical aerated concrete has a density of 600 kg/m3 , and it has other physical
properties comparable to a foam concrete or a light-aggregate concrete of the same
density.

2.2.4 Pervious Concrete

You can design a concrete, which has a system of connected cavities between the
aggregates with a void content typically close to 20%. Such concrete can drain water
efficiently and hinder damages from repeating frost and thaw.
These materials are widely applied for roads in USA, and Iowa State University is
a centre for development of pervious concrete. In cooperation with that university a
PhD project made by Mia Lund [20, 21] developed a pervious concrete, which could
be cast out in a height of more than 2 m, so that you can apply it as a draining and
shear transferring material between arch and top slab in pearl-chain sandwich arch
bridges. See 8.3.2. This material was used and tested in practise in the bridge shown
in Fig. 8.17.
Typical material properties obtained for pervious concrete are a compressive
strength of 10 MPa, E-modulus 16 GPa, Tensile strength 1.8 MPa, and a cavity
content of 17% leading to a density of about 1900 kg/m3 .

2.3 Reinforcement

2.3.1 Slack Steel Bars

A steel material consists of grains with steel crystals. In the crystals, you find some
impurities called dislocations, where a crystal line does not proceed all way through
the lattice in the grain. When you load a bar of mild steel, the crystals at first deform
elastically and the deformation disappears, when you unload the bar. This elastic
deformation gives a stress σ that is proportional to the strain ε as σ = Es0 ε, where
Es0 is the initial modulus of elasticity.
At a certain stress fs the dislocations in the crystal move, and the steel does not
regain its original size, when you unload it. The stress–strain curve will be as shown
to the left in Fig. 2.7. We call this well-defined process yielding and we call this stress
the yield stress. If the steel is mild steel, the strain is about 0.1% and the yield stress
is about 240 MPa. The stress will fluctuate at this level for a while if you increase
the strain. During this process, you create more dislocations in the crystal.
At a certain strain, the stress increases again, because you have created so many
dislocations that they hinder each other’s movement in the steel crystal. We call this
phenomenon cold hardening or strain hardening. If you unload the bar, it will have
a permanent deformation, and if you load it again, this cold-hardened steel will still
2.3 Reinforcement 29

Fig. 2.7 Stress–strain curve for mild steel (left) and hardened reinforcing steel (right)

have the many new dislocations and it will not have a clear yield point. Instead, it
will have a stress–strain curve as the one shown for hardened steel to the right in
Fig. 2.7.
For such hardened steel, you define the “yield stress” fs as the 0.2% stress by
drawing a straight line with the inclination Es0 from the point 0.2% at the strain axis
to the point, where it intersects the stress–strain curve of the cold hardened steel.
From this point, you can consider the steel to deform irreversibly, and you apply the
0.2% stress as a yield stress for design of most steel structures. You can continue to
increase the stress beyond that level contributing to a further cold hardening until the
steel fails and you have reached the ultimate stress fsu .
Slack reinforcing bars were at first plain bars of mild steel with a yield strength
of about 240 MPa and a modulus of elasticity Ec0 = 210 GPa.
However, soon designers of reinforced concrete structures learned to apply cold
hardened bars for the main forces with yield strength or 0.2% strength of about
550 MPa and the same initial modulus of elasticity. Often, the deformation was a
combination of tension and torsion with the purpose of creating more dislocations.
They also provided these bars with a corrugation to ensure a better grip between steel
and concrete. We now only use mild steel for secondary reinforcement.
Instead of making cold hardened bars, industry began to improve strength of
corrugated or deformed reinforcing bars chemically by adding carbon and later other
material atoms to the crystal, which can anchor the dislocations. Deformed bars with
carbon atoms have yield strengths of about 420 MPa. Today, you mainly apply slack
deformed reinforcing bars with 0.2% yield strength of 550 MPa and an initial modulus
of elasticity as Ec0 = 200 GPa. This quality is mainly applied for deformed slack
reinforcing bars in this book.
When you heat up a steel bar, the yield strength or 0.2% strength decreases to about
the half at 500 °C because you need less mechanical stress to move the dislocations in
the crystals [22]. For the same reason, you can create more dislocations by tensioning
the steel further, and within the first 400 °C you can experience an increase of the
cold hardened strength at large strains.
30 2 Materials

This is the reason why you can apply a smaller strength reduction for heated steel
subjected to at least 2.0% strain than the reduction of the yield strength or 0.2%
strength.
Table 2.2 therefore distinguishes between the reduction of 0.2% strength or yield,
which should be applied for most steel structures and slack reinforcement. You can
only apply the reduction of the 2.0% strength, if you can prove that you have a strain
of at least 2.0%. [7].
You can e.g. find such strain of more than 2.0% for plastic design of prestressed
reinforcement and of slack reinforcement in T-shaped beams, where the compression
zone has a small depth compared to the internal lever arm to the reinforcement in
the tension zone.
You can never find a strain of 2.0% in columns and usually not in ordinary slack
reinforced beams or decks, because the deflections should then be so large that the
components would fail due to them. Steel does not become more porous when heated
as concrete, and the stress–strain curves are reduced by an affinity in the strain axis.
This means that

fs (T) = ξs (T)fs20 , Es0 (T) = ξs (T)Es020 and εsu (T) = εsu20 > 2.0%

Any point (εs20 ,σs20 ) at the stress–strain curve is transferred to (εs20 , ξs (T)·σs20 ).

2.3.2 Prestressing Steel

You mainly produce steel for prestressing strands by cold hardening, where large
tensile strains provide it with a high ultimate stress. Due to the high stresses, you
often apply 0.1% yield strength instead of 0.2% strength, so that you apply more
acceptable strains and deformations for plastic design.
For most prestressed structures in this book, we apply prestressing steel with
0.1% yield strength of 1634 MPa, ultimate tensile strength of 1860 MPa, and initial
E-modulus of 195 GPa. For pretensioned SL-slabs and other minor pretensioned
structures, we apply single strands (of seven wires) of a diameter of 12.5 mm with a
nominal steel area of 93 mm2 .
They are often pretensioned to about 72% of the 0.1% yield strength or 1176 MPa.
For larger pretensioned structures and for post-tensioned structures, we mainly
apply strands of 16 mm with a nominal steel area of 150 mm2 .
Usually, you fix strands by wedges in the ends of a prestressing bed.
Factories anchor the straight pretensioned strands with wedges to large steel plates
on heavy concrete blocks. The blocks are placed at each end of the prestressing bed,
which is a mould of often about 100 m in length.
For in-situ cast structures or curved structures, you apply post-tensioned rein-
forcement.
Post-tensioning consist of cables positioned through an embedded post-tensioning
duct. The duct can follow almost any desired shape. A cable consists of a number of
2.3 Reinforcement 31

strands, and the cable duct is often grouted with a mortar after tensioning the cable.
At the ends, the cable is anchored to the concrete. Here, you cast special anchor
blocks and reinforcement into the structure, to fix wedges for the number of strands
in the cable. Usually, you place a splitting reinforcement at the end of the structure
as well distributing the concentrated load of the anchorage to the full cross-section.
For straight prestressing steels, you can alternatively apply solid rods. They are
provided with screw threads and nuts in the ends and may typically have a nominal
diameter between 20 and 75 mm, ultimate tensile strength 1030 MPa, minimum
0.1% proof stress 835 MPa, and modulus of elasticity 185 GPa as seen for example
in [23].

2.3.3 Slack Carbon Fibre Reinforcement

Designers and contractors apply still more carbon fibre reinforcement in new struc-
tures, and for repair of existing structures, where fire load is not important such as
bridges and some industrial buildings.
Carbon fibre reinforcement has very high tensile strength and a modulus of
elasticity comparable to steel. The density is only about 20% of that of steel.
Carbon fibre reinforcement does not suffer from rust as steel reinforcement does.
Designers can therefore apply it with the benefit of smaller or no cover thickness
in many structures. An example of the use of carbon fibre in concrete hinges is shown
in Sect. 9.3.2.
Usually, producers place a number of carbon fibres in a reinforcing bar or a carbon
fibre texture in a flat band. In both cases, they embed the carbon fibres in a polymer.
This is the main reason for the well-known lack of fire resistance. At the author’s lab,
we have made several fire tests of carbon fibre reinforcement and concrete structures
with that reinforcement for example in a project made by Beck and Tvermose [24].
When we heat a carbon fibre bar, the polymer melts and evaporates at temperatures
above approximately 200 °C. We can best describe the result as a horsetail of single
carbon fibres, which seem to have a better fire-resistance than the polymer. We
therefore see a potential for developing carbon fibre reinforcement with a better fire
resistance.
Until that is developed, we have to protect the carbon fibre reinforcement from
fire.
Here, designers can benefit from the application of light and well insulating
concrete in the super-light structures keeping the temperature of the reinforcement
of load-bearing parts below temperatures, where carbon fibre reinforcement melts
(200 °C) or where high-strength concrete explodes (350 °C, see Sect. 2.1.2). This
opens possibilities of new applications of both materials for creating very light
structures with large spans and small impact on the climate.
32 2 Materials

Typical material values for carbon reinforcing bars are tensile strengths of
2000 MPa or 3000 MPa, with initial E-moduli of 250 GPa or 300 GPa, respec-
tively. The ultimate strains are about 1% and no yielding occurs, so the material must
be considered more brittle than steel reinforcement.

2.3.4 Prestressed Carbon Fibre Reinforcement

Prestressing of carbon fibre reinforcement would open new possibilities for design
of super-light structures as described in Chap. 3 and pearl-chain structures described
in Chap. 7.
As mentioned in Sect. 2.3.3, embedding the carbon fibres in light concrete may
solve the problem of strength loss in case of fire. However, anchorage of prestressed
carbon fibre reinforcement bars was a major problem in the past. However, research
by Jacob Wittrup Schmidt has solved the problem [25, 26].
He applies a new design of an aluminium wedge of approximately double length
of that steel wedge used for anchoring of prestressed steel strands. The system has
been applied for external prestressed reinforcement for repair of a number of highway
bridges, and it has been tested in several other structures. This means that the new
technology of prestressed carbon fibre reinforcement may be considered applicable
in general.

References

1. Desideri P, Nervi PJ Jr, Positano G (1979) Pier Luigi Nervi—A cura di Paolo Desideri, Pier
Luigi Nervi jr, Giuseppe Positano. (In Italian) Zanichelli Editore, Bologna, 215p
2. Olmo C, Chiorino C (2010) Pier Luigi Nervi, Architecture as Challenge. Silvana Editoriale,
Milan, 240p
3. CEN (2008) EN 1992-1-1 Eurocode 2, design of concrete structures, part 1-1. General rules
and rules for buildings. Brussels. 225p
4. DS Danish Standards (1984) DS411, 3rd edn. NP-169-N Code for concrete structures. Danish
Society of Engineers Copenhagen. 98p
5. Ritter W (1899) Die Bauweise Hennebique. (In German). Schweizerische Bauzeitung vol 33,
Heft 7, pp 59–61
6. Hertz KD (2005) Concrete strength for fire safety design. Mag Concr Res 57(8):445–453
7. Hertz KD (2019) Design of fire-resistant concrete structures. ICE Publishing, Thomas Telford
Ltd. ISBN: 9780727764447. London, 254p
8. DTU-Byg (2019) Software. Department of Civil Engineering, Technical University of
Denmark. https://www.byg.dtu.dk/english/research/publications/software
9. ISO (1999) ISO 834-1 Fire-resistance tests—elements of building construction—part 1: general
requirements. International Organization for Standardization, Geneva
10. CEN (2006) EN 1992–1–2 Eurocode 2, Design of Concrete Structures, Part 1–2. General rules
– Structural fire design. Brussels. 97p
11. Bache HH (1981) Densified cement/ultra-fine particle-based materials, Aalborg Portland.
Presented at the second international conference on superplasticizers in concrete, Ottawa
References 33

12. Hertz KD (1982) Eksplosion og reststyrke af varmepåvirket silikabeton (Explosion and

Residual Strength of Fire Exposed Silica Fume Concrete. In Danish.) Report 162. Institute of
Building Design (now Department of Buildings and Energy). Technical University of Denmark,
12 p
13. Hertz KD (1984) Heat-Induced explosion of dense concretes. Report No.166. Institute of
Building Design, Technical University of Denmark. Lyngby. Presented at the CIB W14
conference in London May 1984) CIB W14/84/33 (DK), 20p
14. Hertz KD (1985) Explosion of Silica-fume Concrete. Fire Safety J 8(1):77
15. Hertz KD (1991) Silica fume concretes at elevated temperatures. key note speech at a special
session on the author’s findings at the ACI Spring Convention in Boston.
16. Hertz KD (1992) Danish investigations on silica fume concretes at elevated temperatures. ACI
Mater J 89(4):345–347
17. Hertz KD (2003) Limits of spalling of fire-exposed concrete. Fire Safety J 38(2): 103–116
18. Aercrete (2009) Aercrete systemet. Aercrete Sweden
19. Jensen RL, Vahlgren A (2011) Insulation of super-light structures in Arctic regions. Final
project in Arctic Technology, Technical University of Denmark, In Danish (Isolering af
superlette konstruktioner i Arktis), 64p
20. Lund MSM, Hansen KK, Hertz KD (2016) Experimental investigation of different fill materials
in arch bridges with particular focus on pearl-chain bridges. Constr Build Mater 124:922–936
21. Lund MSM (2016) Durability of materials in pearl-chain bridges. PhD thesis, Report R-341
Department of Civil Engineering, Technical University of Denmark. 157p
22. Hertz KD (2004) Reinforcement data for fire safety design. Mag Conc Res 56(8):453–459
23. Macalloy (2009) Macalloy 1030 Post tensioning system. Macalloy Ltd. Sheffield. 8p
24. Beck PP, Tvermose HH (2004) Fire exposed carbon fibre bands. In Danish (Brandpåvirkede
kulfiberbånd) MSc project, Technical University of Denmark, 120p
25. Schmidt JW (2011) External strengthening of structures with prestressed CFRP tendons. PhD
Thesis, COWI Ltd and Technical University of Denmark. 289p
26. Schmidt JW, Krabbe Sørensen JNO, Hertz KD, Goltermann P, Sas PG (2017) CFRP strength-
ening of RC beams using a ductile anchorage system. In: Proceedings of the eighth international
conference on fibre-reinforced polymer (FRP) composites in civil engineering, pp 344–349
Chapter 3
Super-Light Structures

Abstract The core of Super-light technology is to apply the correct and necessary
materials in the optimal position to create a minimal structure. The chapter deals
with how to utilize knowledge of concrete types in guiding the forces through a
construction. A Direct Engineering approach is introduced to learn how it is possible
to choose where the forces should be.

3.1 Super-Light Theory

3.1.1 Minimal Structures

Usually, engineers would like to design structures for a minimum of material

consumption, cost, construction time, CO2 emission, pollution, and other disadvan-
tages. Figure 3.1 shows an example. It is one of the first super-light building projects
from 2010 designed by Bjarke Ingels Group (BIG) Architects with Werner Sobek as
engineer intended for a Building 324 at the Technical University of Denmark.
Builders and researchers try in general to investigate and develop optimal struc-
tures like the vault (Figs. 1.7, 1.8, and 1.9), the cupola (Figs. 1.13 and 1.14), the arch
bridge (Fig. 1.10), or the suspension bridge as seen at the top of Fig. 3.2.
Suspension bridges are often referred to as minimal structures, because the forces
follow a path that results in a minimum use of material in places, where the amount
of material and its weight matters as for example at the free span of the bridge.
The suspension bridge has tension cables in an optimal shape, guiding the forces
to two towers in compression. However, you should also design the bridge to resist
the horizontal tension forces, which often give rise to some large, expensive and
material consuming foundations.
A theoretical alternative would be to counterbalance the horizontal tension forces
at the ends of the cables with compression in the bridge deck. This is almost never
possible in practise, since the deck is usually designed as a light structure that is
too slender to be able to resist these large compression forces because of the risk of
buckling.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 35

K. D. Hertz and P. Halding., Sustainable Light Concrete Structures, Springer Tracts
in Civil Engineering, https://doi.org/10.1007/978-3-030-80500-5_3
36 3 Super-Light Structures

Fig. 3.1 Project for a super-light structure 2010. (BIG Architects)

Fig. 3.2 Examples of minimal structures

At the bottom of Fig. 3.2, you can see an arch bridge, which is the counterpart
of the suspension bridge, where you have exchanged compression and tension in all
structural members. Here, the curved arches have optimal shapes to lead the load of
the bridge as compression forces to the foundations. If you only apply a single arch,
the compression arch may also need some large, expensive and material consuming
foundations to resist the horizontal lateral part of the arch force.
However, if you supplement the arch with two half arches as shown in the figure,
you get the full counterpart of the suspension bridge, and now the bridge deck can
resist the horizontal forces, because these forces are now in tension and therefore,
they do not give problems with buckling. We consider this type of arch structure and
its application in more details in Sect. 7.2.3.
3.1 Super-Light Theory 37

The schematic drawings on Fig. 3.2 do not show all details of importance. Espe-
cially, they lack structural members transferring the continuous weight of the deck
to the bearing cables of the suspension bridge and to the arches of the arch bridge.
These members are hangers in tension at the suspension bridge. For the arch bridge,
they may be columns in compression as shown in Fig. 7.19 or even a lattice as in
Fig. 7.12.
Designers like Gustave Eiffel (Fig. 7.8) and Edgar Cadoso (Fig. 7.19) have devel-
oped and refined minimal structures for their large bridges, and researchers like [1–3]
have developed general theories for minimal structures. However, when you reduce
the size of the structure, you will find it more difficult to apply a minimal shape for
a design that is optimal with respect to all parameters.
The smaller you make the structure, the closer, you have to place the hangers or
columns to benefit from the curved shape. The connections then occupy a relatively
larger part of the design and of the total resource consumption. The design therefore
becomes more clumsy and less optimal.
In buildings, most dimensions are considerably smaller than in the large bridges,
and it is even more clear that an optimal design requires a holistic approach. An
optimized design depends on fulfilment of a number of very different functional
requirements like sound insulation, fire resistance, sound damping, integrity, heat
insulation in addition to the load-bearing capacity. The shape of an optimal structure
can therefore seldom be the same as the one of a construction, which is minimal with
respect to material consumption for the load-bearing function alone. A few gifted
master builders like Appollodorus from Damascus (Fig. 1.12) and Piere Luigi Nervi,
see Figs. 3.3 and 2.1, have shown us examples where a building shape coincides with
a minimal structure [4, 5].

Fig. 3.3 Palazzetto dello Sport by PL Nervi 1958. (Drawing OSR Frederiksen)
38 3 Super-Light Structures

3.1.2 Super-Light Principle

In 2007, the first author invented super-light structures in order to solve the problem
of an optimal outer shape, that would not necessary be identical with the minimal
shape of the load-bearing structure [6]. You achieve this by designing a primary
minimal load-bearing construction, and place it in a light structure with an outer
shape fulfilling a number of other requirements. The outer shape could for example
be a rectangular beam that provides you with a horizontal surface on top. The light
material, which in this case could be light concrete, transfers the load from the top
of the beam to the embedded primary load-bearing construction, which in this case
could be a curved arch of strong concrete. Furthermore, the light concrete stabilizes
the stronger arch against buckling, so that you can reduce the required size of the
arch cross-section and reduce the amount of strong concrete applied.
You may for example design a primary load-bearing structure as an arch (Fig. 3.4)
of an ordinary 55 MPa concrete of density 2300 kg/m3 and cast an outer structure of
light aggregate concrete of density 600 kg/m3 with a compressive strength of only
3 MPa. The strong arch may require a curved and expensive mould like those, we
applied in the very first super-light test structures as for example the double beam
shown in Fig. 7.10 made by Larsen [7]. The expensive mould has been a main reason
why builders have rarely applied arches and vaults for more than 50 years in most
industrialized countries.
We therefore soon found a more economical solution as shown in Fig. 3.5, where
we hang a tube with the shape of the arch upside down, cast the light concrete around

Fig. 3.4 Super-light principle with a strong arch in a light material

Fig. 3.5 Concrete arch cast in a tube and embedded in light concrete
3.1 Super-Light Theory 39

it, and then cast the strong concrete in the tube. The tube could be a duct similar to
those applied for post-tension cables, but you may apply less expensive solutions,
because the light concrete gives a smaller mould pressure. Later, we invented the
pearl-chain principle [8] as an inexpensive alternative to the curved mould. Chapter 6
explains more about this.

3.1.3 Benefits and Goals

The super-light principle allows you to design a structure with a minimum of material
consumption, CO2 emission, and transport costs and still fulfill the many different
functional requirements.
The application of a composite structure with two different concrete materials
with different densities and elastic moduli gives you another benefit. The concrete
parts oscillate with different eigenfrequencies and the structure therefore reduces
sound energy to heat.
This means that you can obtain a better sound insulation than possible, if you
only applied one material. We utilize this in SL-decks (Chap. 4) and composite
walls (Chap. 5) to reduce material consumption and CO2 emission of structures for
domestic buildings and other places, where sound insulation is required or wanted
(Chap. 11).
The light concrete is very fire resistant [9] and has an insulating effect. This means
that if you embed strong concrete in an outer shape of light concrete, you protect the
strong load-bearing concrete from fire and you may obtain a considerably increased
fire-resistance. We explain this further for the super-light deck elements in Chap. 4.
You may even design your structure so that you keep the temperature of the strong
concrete below 350 ˚C. This means that you can apply a high-strength concrete of
more than 80 MPa to carry the load without a risk of explosive spalling [9].
Explosive spalling has so far been a hindrance for application of high-strength
concrete in buildings and tunnels. However, super-light structures solve this major
problem.
Since light concrete can also stabilize the strong, you can now apply very slender
cross-sections of high-strength concrete. This allows you to design super-light
structures of extremely low weight and for example obtain long span widths.

3.2 Direct Engineering

3.2.1 Direct Engineering Principle

In traditional structural engineering, you assess the outer loads and choose a shape
of the structure according to your experience of how structures usually are.
40 3 Super-Light Structures

Then, the next task is to find out, how the forces are distributed in the structure.
To do this, you make a structural analysis.
This is a rather demanding discipline, where you make qualified presumptions
concerning stress–strain relationship of involved materials, how stresses can be
distributed, how plastic hinges can occur, how cracks may develop and change the
stress distributions, etc.
Finite element programs are often applied to give an idea of how stresses and
forces could be distributed, and the outputs of these programs are approximations
depending on a number of assumptions. Finally, you find a design of the structure
that can contain a distribution of forces, which can resist the loads.
The principle of super-light structures allows you to reverse the engineering
process.
Here, you can decide, where you want the forces to be. This means that you decide
an optimal shape of the load-bearing parts of strong concrete. These may for example
be a minimal structure of arches as the simple example shown in Fig. 3.4 and perhaps
you supplement them with some elements in tension.
Then you fill the outer shape with a light concrete that stabilises and protects the
strong.
Since you know the forces that you want the single elements to transfer, you can
immediately design the strong elements to resist them.
Since the strong elements have a considerably larger stiffness (about 10 times
larger modulus of elasticity) than the light concrete, you can be sure that each force
will find the place, where you have decided it to be. This means that you are not
required to make a structural analysis.
Especially not, if you consider that you have a lower bound solution. (See
Sect. 3.2.2).
Of course, you may do so in order to obtain an extra check of your structure.
This way of working with construction design appears to be straight forward, and
we therefore call it “Direct engineering” in contrast to the traditional that appears to
be backwards and indirect.
You may say that in direct engineering, you decide how the structure should
perform instead of accepting that the structure determines how you should perform
(Fig. 3.6).

Fig. 3.6 Direct engineering

3.2 Direct Engineering 41

When you design super-light structures with direct engineering, it is sometimes

helpful to consider the structure as a skeleton of strong concrete embedded in a body
of soft parts of light concrete perhaps interacting with sinews of tension members
such as reinforcement.
The analogy to the body anatomy can inspire you to find new solutions for
structural problems.
As mentioned, you can make the curved shapes of the compression zones inex-
pensively and effectively as pearl-chains according to Chap. 6. It is possible to
design pearl-chains of mass-produced straight elements, which constitute the curved
structures like spines in a skeleton analogy.
Tension members may be ordinary steel reinforcement bars and cables, but they
may also be carbon fibre rods or lines. In many cases, this is a new possibility, because
the light concrete can provide a fire-protection required for application of such mate-
rials. Carbon fibre does not corrode. This therefore opens new possibilities of placing
tension reinforcement in light concrete exposed to a wet climate or reducing concrete
cover layers.
You may also design tension members as prestressed concrete elements. They can
be mass-produced as pre-tensioned structures by an element factory. They can also
be designed as post-tensioned concrete elements with a cable duct, and in this way,
you can even create curved pearl-chain tension ties of prestressed concrete.
The benefit of these prestressed concrete structures is that they can resist tension
as unloading of the pre-compression in the concrete. This has the advantage of a
considerably larger axial stiffness, and consequently a smaller deflection and higher
loads before cracking of the super-light structure.
Like carbon fibre reinforcement, you can apply prestressed concrete tension
members embedded in a light concrete without problems of corrosion.

3.2.2 Lower Bound Solutions

If you can find a distribution of forces that can sustain the load on the structure, it is
statically possible. If the structure can sustain the forces in any point, it is safe.
A statically possible and safe force distribution is per definition a lower bound
solution.
If the internal force distribution in the physical structure is not as postulated, the
structure is either able to sustain the applied load, or the forces may redistribute
through cracking and plastic deformation into new positions for example as the
postulated, where they can sustain the load. Since the postulated force distribution is
proven to be possible and safe, the structure at least has one possible way of resisting
the load, and it can therefore not collapse. This argumentation gives the third criterion
that the internal force distribution should be able to change from the actual one in
the physical structure, to the postulated one in the calculation model. This is often
referred to as the structure having a sufficient rotational capacity.
42 3 Super-Light Structures

No matter how you calculate a force distribution, you have to make simplifications
and presumptions such as application of elastic-, plastic-, or other calculation models,
a crack distribution model, a material model etc. This means that you actually always
calculate a system that is different from the physical structure.
Therefore, if you find that your force distribution is safe, the calculation will
be a lower bound solution. It is lower bound, because there may be another force
distribution giving a larger resistance to the load. If that is the case, this other force
distribution will be the one applied of the structure at the ultimate limit state. These
considerations give the following conclusion:
At the ultimate limit the forces and stresses in a structure will be distributed so,
that they give maximum resistance to the load, and that any other postulated safe
and statically possible distribution will be a lower bound solution provided that the
structure can redistribute forces from the postulated to the ultimate.
The principle of lower bound solutions has led to simple design methods such as
the strip method introduced by Hillerborg [10], where a slab is considered to consist
of two crossing sets of strips, where you postulate that the strips in one direction
resist a certain part of the load, and de others the residual load.
Other researchers, as for example by Nielsen and Hoang [11], have developed
theories about lower bound solutions further.
In super-light structures, you create a core or “skeleton” of strong concrete, where
you postulate that the compression forces will be.
Perhaps you also have the skeleton interacting with some tension members and
then you fill the outer shape with a light concrete material.
If the skeleton can resist the loads when the light concrete stabilizes it, you have
a lower bound solution, and you have designed the structure.

References

1. Andersen S (1979) The accumulated energy consumption for building materials. Report 134,
(in Danish) DTU Byg (Inst.of Building Design) 1979, 198p.
2. Andersen S (1980) The accumulated energy consumption for dwellings. Report 137, (in Danish)
DTU Byg (Inst.of Building Design) 1980, 124p.
3. Reitzel E (1979) From fracture to shape. (in Danish), Polyteknisk Forlag, 269p
4. Desideri P, Nervi PJ Jr, Positano G (1979) Pier Luigi Nervi—A cura di Paolo Desideri, Pier
Luigi Nervi jr, Giuseppe Positano. (In Italian) Zanichelli Editore, Bologna, 215p.
5. Olmo C, Chiorino C (2010) Pier Luigi Nervi, Architecture as Challenge. Silvana Editoriale,
Milan, 240p
6. Hertz KD (2010) Light-weight load-bearing structures. (Super-light structures). Applica-
tion no. EP 07388085.8. European Patent Office, Munich. November 2007. Application no
61/004278 US Patent and Trademark Office. November 2007. PCT (Patent Cooperation Treaty)
Application PCT/EP2008/066013, 21. November 2008. Patent 2010.
7. Larsen F (2007) Light Concrete Structures. Department of Civil Engineering, Technical
University of Denmark, MSc Project, p 117
8. Hertz KD (2012) Light-weight load-bearing structures reinforced by core elements made of
segments and a method of casting such structures. (Pearl-Chain Reinforcement). Application
no. EP 08160304.5. European Patent Office, Munich. July 2008. Application no 61/080445
References 43

US Patent and Trademark Office. July 2008. PCT (Patent Cooperation Treaty) Application
PCT/EP2009/052987, 13. Mar 2009. Patent 2012.
9. Hertz KD (2019) Design of fire-resistant concrete structures. ICE Publishing, Thomas Telford
Ltd. ISBN: 9780727764447. London, 254p
10. Hillerborg A (1975) Strip method of design. Cement and Concrete Association. 256p
11. Nielsen MP, Hoang LC (2016) Limit analy CRC Press, Taylor & Francis Group. First published
in 1984:816p
Chapter 4
Slabs and Beams

Abstract The mass produced SL-Deck is a slab, which is based on the tech-
nology of Super-light concrete. It is light-weight, and has many benefits such as
fire safety, sound insulation and flexibility regarding geometry and connections. The
SL-Decks can be used in ordinary buildings, in beamless buildings, and for many
other applications.

4.1 SL-Decks

4.1.1 Prototypes of Super-Light Deck Elements

When we invented the super-light structures described in Chap. 3, we could see

a wealth of interesting applications, where builders could benefit from the new
technology.
It was soon realized that producers of the first components would like to have
a market of a certain size, where the products meant a considerable improvement
in terms of a lower price for a better performance. Prefabricated deck-elements
represented such a market.
In cooperation with BIG Architects, their applications were investigated in
advanced building design [1]. By creating a mass-produced super-light deck element,
material consumption and CO2 emission could be reduced and fire-resistance and
flexibility improved. Odd geometries of decks could be created with large holes, and
embed services such as electricity, floor heating, water supply, and sewers in the
elements.
In addition, we could prepare transfer of moments for vertical load in continuous
slabs and shear for horizontal load to other structural components, where the slabs
act as horizontal shear plates. Fixings could also be implemented for example for
balconies and for preliminary handrails needed at the building site.
We therefore started to design a super-light deck element called the SL-deck.
Soon, it was realized that it should consist of some blocks of light-aggregate
concrete, which could serve as permanent moulds for a stronger concrete.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 45

K. D. Hertz and P. Halding., Sustainable Light Concrete Structures, Springer Tracts
in Civil Engineering, https://doi.org/10.1007/978-3-030-80500-5_4
46 4 Slabs and Beams

The strong concrete can carry the load in small arches across the blocks. The
arches rest on reinforced ribs in two directions between the blocks. This means, that
the strong concrete constitute a ribbed structure with vaulted plates between the ribs.
At that time in 2011, concrete factories were not able to cast strong- and light
concrete together. We therefore asked a light aggregate concrete factory to assist us
casting the first blocks. In cooperation with them, we found a shape for the blocks,
which a machine could easily cast, and which could adhere to a strong concrete.
A series of blocks was made and transported to a concrete element factory that
produced a number of prototype deck elements. Figure 4.1 shows the precast blocks
placed on a 1.2 m wide and 100 m long track. The structure was reinforced with
pre-tensioned 12.5 mm tendons along the track, and slack deformed (corrugated)
bars across. Then we cast the strong concrete on top of this. Next day, we could lift
the super-light deck elements off the track.
Comprehensive tests were made of the prototype elements (Fig. 4.2) for bending
[2], shear [3], and anchorage [4]. In addition, we tested them for fire resistance.
We considered the load case of fire very important, because new requirements for
heat insulation have increased fire exposure representing a doubling of the required
standard fire resistance [5]. See more about this subject in Sect. 4.1.4 and in Hertz
[6].
Finally, a large effort was made for designing and testing the prototype elements
for sound insulation in cooperation with the acoustical department of the consulting
engineering company Grontmij [7]. Usually, the requirement of a proper sound insu-
lation gives rise to design of heavy decks and walls and hereby a larger production
of CO2 . See more about this subject in Sect.4.1.3.

Fig. 4.1 Blocks for 1.2 m wide prototype SL-decks in 2011. Photo KD Hertz
4.1 SL-Decks 47

Fig. 4.2 Bending test of prototype SL-deck. Photo KD Hertz

The Danish Building Regulations has for example a minimum requirement of

55 dB sound insulation. This is a main hindrance for design of light building structures
saving materials.
The Technical University of Denmark applied a number of the prototype deck-
elements for balconies in a new building 324. The architect Christensen and Co.
designed balconies as hung-down deck structures in two levels (Fig. 4.3). Ordinary

Fig. 4.3 Hung down balconies of prototype SL-decks. Photo KD Hertz

48 4 Slabs and Beams

deck elements could not adopt the concentrated forces from the bearing rods, but SL-
decks allow incorporation of such bearings. In addition, the balcony decks included
blade connections so that one deck can rest in plane on the next without application
of beams. (See more about that in Sect. 4.2.6).

4.1.2 Mass Produced SL-Decks

Based on the experience from design, test and application of the prototype decks, we
designed the first SL-deck for mass production as shown in Fig. 4.4. The width was
assessed to be 2.4 m, which is the double of that of the prototype and maximum of
other prefabricated deck elements. We introduced a wider groove between the light-
aggregate concrete blocks at the middle of the cross-section allowing implementation
of special services or a duct for a post-tensioning cable. Figure 6.7 shows such
elements, where the cable duct is visible in a recess, so that you can connect it to the
duct of the next element before casting mortar in the joint.
This means that we have prepared the SL-deck for application as a inexpensive,
mass-produced element for pearl-chain structures. The SL-deck is therefore not only
intended to be used as slabs in buildings but also as components for arches and vaults
and for bridge design as explained in Chaps. 6 and 7.
Figure 4.5 shows 100 m long and 2.4 m wide tracks for casting SL-decks, In the
prestressing bed, magnetic mould steel plates separate the decks. The sides can turn
and automatically de-mould the elements. We shaped the sides in the longitudinal
joints as shown in Fig. 4.6 so that it can transfer horizontal and vertical shear to
the next element. The profile allows pouring of mortar in the joint without a shutter

Fig. 4.4 2.4 m wide SL-decks

4.1 SL-Decks 49

Fig. 4.5 2.4 m wide tracks for SL-deck production. Photo KD Hertz

Fig. 4.6 Detailing of the longitudinal sides of a SL-deck. Photo KD Hertz

board beneath. Special bearing knobs in the sides of the elements serve to lift the
longitudinal reinforcing bar in the joint in order to embed it in the mortar ensuring
its bond to the deck and to protect it against moisture and corrosion from the bottom
of the joint.
At first, the factories usually place a thin layer of mortar on the mould, upon
which the blocks are placed. This ensures a smooth underside of the elements, which
customers usually want for their ceilings in domestic buildings.
50 4 Slabs and Beams

Fig. 4.7 Automatic block casting machine for SL-decks. Photo KD Hertz

If wanted, the factory can keep the sides free from this bottom mortar layer to that
there are voids along the element sides. This allows the contractor to fill the voids at
the building site and thereby obtain a smooth unbroken ceiling without visible joints
between the elements.
Figure 4.7 shows an automatic machine developed to cast the light-aggregate
blocks. It can place blocks, where you have programmed it on a 100 m long track in
two hours.
This means that you can omit blocks and hereby create spaces in the deck, where
you can have strong concrete, holes, services, or inlayed structural elements such as
bearings, beams etc.
Figure 4.11 shows an element with an inlayed beam at the end for lateral support.
Figure 6.17 shows an advanced example of a post-tensioned inlayed beam
designed for bending, shear and torsion.
The strong concrete applied has usually a compressive strength of 55 MPa, and the
density of the light-aggregate concrete is about 700 kg/m3 . The machine casting the
light aggregate concrete blocks has a row of six moulds (for 2.4 m SL-deck widths).
The number of moulds used in each row and their position can be controlled.
The machine vibrates the light-aggregate concrete, lifts the moulds, and proceeds to
a new row. Then, the factory workers places tendons, which are usually 12.5 mm,
along the track and pretension them. Then slack reinforcement, inserts and services
etc. are positioned and finally, the strong self-compacting concrete is cast. The strong
concrete penetrates the light before the light hardens and creates a zone of a couple
of centimetres, where the quality gradually changes from the one to the other without
forming a surface separating the two materials. Even at full-scale failure tests, the
light blocks do not separate from the SL-deck.
4.1 SL-Decks 51

By combining pre-programmed placing of blocks and placing of magnetic fixed

mould pieces, you can create deck elements of any shape, you want. Figure 4.8 shows
an irregular element lifted in place at a building site in Copenhagen Harbour.
Figure 4.9 shows a SL-deck element that demonstrates possibilities of making
holes in the decks. The large rectangular holes were for skylights, and the engineer
placed inlayed reinforced beams across the element between them. The element is
still so freshly produced, that you can see the position of the light-aggregate blocks

Fig. 4.8 Irregular SL-deck element. Photo Abeo Ltd

Fig. 4.9 SL-deck with holes spanning over several bays. Photo KD Hertz
52 4 Slabs and Beams

Fig. 4.10 Wedge shaped SL-deck elements for a round building. Photo Abeo Ltd.

as a pattern at the bottom caused by differences in moisture. The element is also an

example of a continuous SL-deck designed to span over a number of rooms. This is
obtained by placing top-reinforcement so that negative bending over the intermediate
supports can be resisted.
By utilizing the option of continuous decks, you can make fewer crane lifts at the
building site and transport fewer elements. You may also increase the possible span
lengths by designing elements with fixed end supports.
Figure 4.10 shows how a series of equal wedge shaped SL-deck elements with
rounded ends constitutes a floor structure of a round building. The picture also
demonstrates how you can place preliminary safety handrails using inserts in the
deck prepared for the purpose at the factory. This saved the workers 1–2 days of
drilling and applying fixings for the safety handrails at the building site, which they
would normally do for other deck solutions.
SL-decks are produced in thicknesses typically varying from 180 mm over
220 mm, 270 mm, 320 mm, and up. The strong concrete has typically a minimum
thickness of 40 mm above the light blocks. Either you can choose to increase the
block height as the total deck height increases, or you can keep the blocks height
and just increase the total deck height. If you choose to increase the block height,
it is done by varying the height of the inclined sides of the blocks, where all other
measures of the block geometry remain constant.
The pretensioned tendons are usually 12.5 mm in diameter with a cross-sectional
area of 93 mm2 . Two of them are placed at the sides of the element with 42 mm
from the bottom to the centre line, and the others in the grooves between the blocks
a levels of 60 and 92 mm from the bottom to the centre lines. 20 mm of this height
from the bottom is insulating light-aggregate concrete.
4.1 SL-Decks 53

Fig. 4.11 SL-deck element with inlayed beam for lateral support. Photo KD Hertz

Top reinforcement tendons are typically placed 30 mm from the top surface of
the element to the centre lines.
At each end, you find a zone of massive concrete with a length of at least 200 mm,
where the pretensioned reinforcement is anchored, and where the deck element is
supported. This means that the presence of deck ends in joints between walls and
decks does not introduce weak areas affecting the strength of a wall in a multi-storey
building. As shown in Sect. 9.1, this may be more than sufficient for anchorage of
the tendons because the compressive stresses from the bearing increases the splitting
strength at maximum load.
As mentioned, the first Danish factory produces SL-decks of width 2.4 m. The
cross-section of these elements has a maximum of six light-aggregate blocks of
width 375 mm as shown in Fig. 4.12. This width is the maximum obtainable for other
mass-produced concrete deck elements without a cross-reinforcement. However, SL-
decks have a cross-reinforcement and they do not have this limit in order to avoid
longitudinal rupture. The limit of the width is therefore solely determined by the
width that can be transported by trucks without special permission. This is found to
be 3.0 m, which has therefore become the new standard width for European factories.
The cross-section of these elements has a maximum of eight light-concrete blocks
each of width 350 mm as seen in Fig. 4.12 leaving two wider grooves, where ducts
for post-tensioning cables can be placed for application in pearl-chains (Chap. 6).
In USA, two factories are built and more are on the drawing board.
Here the SL-deck is called T-slab, and the standard width is 12’ or 3.6 m. This
large width suites well to the American standards of building and allows very few
expensive crane lifts.
54 4 Slabs and Beams

Fig. 4.12 Produced SL-deck cross-sections, 2,4 m and 3.0 m in Europe, 3.6 m in USA

4.1.3 Fire Resistance of SL-Decks

As mentioned in Sect. 4.1.1, we considered fire safety of deck constructions

important.
Since 1990, building regulations prescribed improvement of building insulation
in order to save energy. The result is, that a larger part of the heat from a fire remains
inside, and that windows seldom break. This reduces ventilation and increases fire
duration and temperatures in structures. You consider this, when you design a struc-
ture for a fully developed fire. However, you often make an alternative design for a
fixed required fire-resistance time of a standard fire curve without a cooling phase
according to ISO 834 [8].
The authorities have not yet adjusted these fixed requirements in the building
regulations according to the better insulation of the buildings. If you for example
compare a temperature of a reinforcing bar, you can see, that you should double the
required standard fire-resistance time, if a standard fire design should still be relevant
for the real fully developed fire in a modern building [6]. We therefore designed the
SL-deck elements for a 240 min standard fire-resistance, so that it can withstand the
same fires, as we previously represented by a 120 min requirement.
We first tested two prototype slabs before we made an accredited test on a mass-
produced 220 mm thick, 2.4 m wide, and 6 m long SL-deck (Fig. 4.13).
After 240 min standard fire exposure, the bottom of the deck was 1154 °C, the
prestressed reinforcement was 389 °C, and the top of the deck was 58 °C. The deck
was unharmed and the thermal deflection disappeared, when the deck cooled down.

4.1.4 Sound Insulation of SL-Decks

Sound insulation means reduction of noise from one room to the next across a sepa-
rating deck or wall. Deck structures are required to have a reasonable sound insulation
4.1 SL-Decks 55

Fig. 4.13 SL-deck element after 240 min standard fire test. Photo KD Hertz

if users of a building should not disturb one another. In most countries, the building
regulations require a minimum sound insulation, and in Denmark, it is 55 dB. Sound
insulation requirements are major hindrances for making light sustainable structures,
because you need mass in order to fulfil them. If a deck is made of massive concrete,
you must have 440 kg mass per m2 to obtain a 55 dB sound insulation.
A SL-deck consists of heavy and light concrete. The two materials oscillate differ-
ently, and the deck therefore reduces some of the noise to heat. This is the reason
why, the SL-deck can obtain more than 55 dB sound insulation with a weight of only
350 kg/m2 .
We tested prototype decks [7] and made accredited tests of the mass-produced
220 mm SL-deck.
It documented that a bare deck element had a sound insulation of 57 dB, and
installed with a standard floor, the sound insulation was 58 dB, where the requirement
is 55 dB. (Fig. 4.14).
Step noise or impact sound level was measured with a standard impact machine
to be 47 dB, which is 6 dB better than required.

4.1.5 Sound Damping of SL-Decks

Where sound insulation is reduction of noise transfer between rooms, sound damping
or sound absorption is reduction of noise in the room itself. Bare concrete surfaces
may reflect sound and give rise to a hard acoustic (strong reverberation) in buildings.
Often, builders apply wood-wool claddings in order to damp the sound. The porous
56 4 Slabs and Beams

Fig. 4.14 SL-deck element tested for sound insulation and step noise. Photo Abeo Ltd.

Table 4.1 Sound absorption of light-aggregate concrete compared to wood-wool

Frequency in Hz 125 250 500 1000 2000 4000
Light-aggregate 0.15 0.19 0.44 0.63 0.52 0.75
Wood-wool 0.08 0.20 0.45 0.80 0.66 0.85

surface of the wood-wool absorbs the sound energy by means of a kind of friction,
when the air oscillates in the small cavities of the material.
When you apply SL-decks, you can have a similar sound damping by omitting the
bottom layer of mortar, which the factory usually place in order to deliver a smooth
surface.
If you omit that, you will have a free porous surface of the light-aggregate
concrete that can provide a sound damping effect. Table 4.1 shows that the sound
damping even as a function of frequency is almost identical to the one of wood-wool
slabs. Figure 4.15 shows a classroom at Gammel Hellerup Highschool, where BIG
Architects utilized the sound damping of the SL-decks in the ceiling.

4.2 Super-Light Deck Details

4.2.1 Fixed End Connections

If you design a continuous deck element supported by e.g. a number of walls, you
can obtain longer span widths and save material and CO2 . Utilizing fixed ends is a
4.2 Super-Light Deck Details 57

Fig. 4.15 Classroom with sound damping SL-decks. Photo KD Hertz

way to obtain longer spans and save CO2 as well. Moreover, both design options
save crane lifts and mounting time at the building site. SL-decks like the one shown
in Fig. 4.9 are therefore often designed with several continuous spans, and provided
with a top reinforcement in order to resist the negative moments at the supports.
If shear is a problem near a supporting wall, you may design a massive zone by
omitting to place some blocks and reinforce it with stirrups as a beam.
You may also provide the deck element with holes to connect vertical wall
reinforcement.
However, the length that a truck can carry in ordinary traffic and the weight that the
applied cranes can lift, represent limits for the length of the deck elements. You may
therefore wish to create a joint, where you can make a fixed end connection between
two deck elements over a supporting wall. Figure 4.16 shows a simple design of such
a joint. Slack top reinforcement is placed at the ends of the SL-decks, where you can
expect negative moments and this slack reinforcement is connected to steel plates,
which can be simply connected with bolts to the similar plates in the other element.
The holes required for inserting the bolts are filled with mortar at the same time as
the rest of the joints.
Thinking of the disassembly process of the building, a weak mortar can be used in
the hole, so that the bolt can be exposed again for easy dismantling. (This is explained
in more detail later in Sect. 9.2.1).
Alternatively, you can provide the deck elements with grooves, where the
contractor at the building site can place a top reinforcement over the joint at the
wall.
58 4 Slabs and Beams

Fig. 4.16 Fixed end connection between SL-decks. Photo Abeo Ltd.

4.2.2 Connection to Balcony

Fixed end connections are also required for cantilevered balconies. Figure 4.17 shows
an example, where top reinforcement and inclined shear reinforcement are concen-
trated in two prefabricated zones of low conductivity between a SL-deck and a
balcony.

Fig. 4.17 Fixed end insulated connection to balcony. Abeo Ltd.

4.2 Super-Light Deck Details 59

4.2.3 Connection to Column

If a column should support a SL-deck element without application of a beam, you can
create a zone of strong concrete above and around the column by omitting blocks in
the area. The zone of strong concrete may also serve to transfer forces from a column
above the deck to the column below, and holes can be made in the deck in order to
connect vertical reinforcement combining the column elements. The deck usually
has to be supplied with horizontal reinforcement in two directions in order to resist
negative moments. Probably, it will also be necessary to insert stirrups or inclined
shear reinforcement in the deck around the column connection.

4.2.4 Safety Handrails

As shown in Fig. 4.10, the contractor can place safety handrails in fixings cast into
the deck at the factory, and usually, the elements are lifted in place with the handrails
mounted.
This saves time at the erection site for drilling fixings and mounting the handrails.

4.2.5 Services in SL-Decks

Sockets for vertical pipes and tubes for horizontal pipes and electrical cables can be
cast into deck elements at the factory. They can be provided with packings to allow
a quick connection at the building site.
Figure 4.18 shows an example of a bathroom floor prepared with drain, water
supply, floor heating, and electricity as services built in at the factory. This saves
time at the building site. It also save height, because you do not have to place a
bathroom floor or bath unit on the load-bearing deck element or alternatively to
place services into an in-situ cast top concrete layer.
Often, floor heating is placed as a net of pipes cast into the top of the SL-deck
at the factory, where the thickness above the blocks is adjusted for the purpose. The
pipes are connected to those of neighbour elements via couplers in recesses at the
edge of the element.

4.2.6 Blade Connections

As mentioned in Sect. 4.1.1 and shown in Fig. 4.3, we developed blade connections
already for the prototype version of the SL-deck. A blade connection is a bearing
60 4 Slabs and Beams

Fig. 4.18 SL-deck with built-in services. Abeo Ltd.

detail, where one element can rest on another in the same plane without application
of a beam or any other increase of the structural thickness.
A blade connection is created in a zone of strong concrete of a deck element by
halving the slab thickness in a length required for the bearing.
It is necessary to provide the half thickness zone with a shear reinforcement
for example a set of stirrups or a number of inclined reinforcing bars as shown in
Fig. 4.22. The horizontal parts of the reinforcing bars near the bearing plane serve
also to resist tension from bending of the blades.
Figure 4.19 shows a blade connection, where the end of a SL-deck element is
supported on a blade inlayed in the side of another deck element. In case the last
element is a SL-deck, you can omit a number of blocks, in order to make space for
the blade at the side and for the reinforcement required to distribute the reaction force
into the deck element.

4.3 Beamless Super-Light Decks

4.3.1 Principle of Beamless Decks

The blade connection presented in Sect. 4.2.6 allows us to establish a bearing of

a slab at an end or at a side of another slab without giving rise to any increase of
thickness, and without any beams. We apply this connection in the beamless deck.
We support the deck on columns, and as mentioned in Sect. 4.2.3, we need a zone
of strong, reinforced concrete in a deck above a column. This is required, if it should
4.3 Beamless Super-Light Decks 61

Fig. 4.19 Blade connection of SL-decks. Photo KD Hertz

transfer load from a column above to a column below the deck. As also mentioned
some top reinforcement is required in two directions to resist negative moments, and
perhaps some stirrups or inclined reinforcement to distribute the column reaction
into the deck.
We propose using a special quadratic massive slab element with blade connections
at all four sides and a width equal to the width of the SL-decks (could be 3.0 m). We
place such quadratic slab element at the top of each column as shown in Fig. 4.20.

Fig. 4.20 Beamless super-light deck

62 4 Slabs and Beams

Between quadratic top slabs above adjacent columns, we place a main deck element
with blade connection bearings like that in Fig. 4.22 along the sides. Figure 4.20
shows this with a darker colour.

4.3.2 Moment Distribution in Beamless Decks

A number of SL-decks are supported by the blade connections at each side of the
main deck element. The SL-decks are reinforced in the bottom for positive moments,
and since you have zero moment at the blade connection, the main deck element is
provided with a top reinforcement for the negative moments shown to the left in
Fig. 4.21 in the direction across its length.
The main deck elements can be made of massive concrete, and they have bottom
reinforcement in their length direction, because they only have to resist positive
moments in that direction. This is due to the moment being zero in the connection to
the column deck element. With such structure, you can e.g. create a beamless office
building or parking garage [9]. It is an advantage that you save height, and the negative
moment zones prolong the span width of decks without increasing their thickness.
At the sides of the building, you may create wider top slabs to counterbalance the
negative moment from the one side.

Fig. 4.21 Moment distribution in a beamless deck

Fig. 4.22 Blade connection with shear reinforcement

4.3 Beamless Super-Light Decks 63

4.3.3 Design for Disassembly

The blade connections are in most cases required to be poured with a mortar to ensure
the building robustness. These mortar joints are only half as deep as the deck height,
and this is an advantage if you should consider the disassembly and reuse of the
elements. If a weak lime mortar is used, then it will be easy to remove the mortar
and dismantle the building at the end of life [10].

4.4 Light Concrete Slabs

4.4.1 Massive Light Concrete Slabs

Builders have applied light aggregate concrete for slabs since antiquity.
For in-door structures, where reinforcement is not exposed to moisture, reinforced
massive slabs of light aggregate concrete are quite common.
The first author has investigated fire-resistance of a number of such slab elements
[11], which are representative for the production. A typical mass-produced massive
deck element could be 200 mm thick, 1.2 m wide and 6 m long. It is made of a
light aggregate concrete of density 1775 kg/m3 with a compressive strength of fcc20
= 20 MPa, tensile strength fct20 = 3.2 MPa and conductivity λ = 0.6 W/mK. It is
reinforced with 8 Y 10 mm slack deformed bars of tensile strength 550 MPa with a
cover thickness d = 15 mm.
The weight of the slab is about 360 kg/m2 .
The anchorage length of the deformed bars at the ends of the slab is 70 mm. We
could prove by calculation as well as by test that this is sufficient according to the
explanations given in Sect. 9.1.
We calculated the standard fire resistance for bending, shear and anchorage for
this slab to be 69 min, and proved by test that it was 73 min for a service load of
2.4 kN/m2 (Fig. 4.23).
Limiting parameters for designing deck structures of light aggregate concrete
can be span-width, shear, anchorage, and for domestic buildings sound insulation
requirements.
A density of about 1800 kg/m3 does represent a 20% reduction of dead load
compared to a similar massive slab of ordinary concrete, and this may have a little
positive influence on the possible span-width. However, it has a negative influence
on the sound insulation.
If you should obtain a usual minimum requirement of 55 dB, you need 440 kg/m2
and a thickness of 250 mm. You may reduce this if you test the deck in combination
with a floor structure or an in-situ cast top concrete layer.
64 4 Slabs and Beams

Fig. 4.23 Fire test of 200 mm light aggregate concrete slab. Photo KD Hertz

4.4.2 Sandwich Slabs

If you apply a light core and strong flanges, you can obtain a light sandwich slab with
a reasonable load-bearing capacity. You can typically apply such structure, where
sound insulation is not required.
An example of a mass-produced sandwich deck element is 240 mm thick, 1.2 m
wide and 6 m long. It is made of.
(1) a 23 mm thick top of light aggregate concrete of density 1550 kg/m3 with
compressive strength fcc20 = 15.25 MPa
(2) a 182 mm thick core of 625 kg/m3 light aggregate concrete with and
compressive strength fcc20 = 2.8 MPa and tensile strength fct20 = 0.3 MPa
(3) a 35 mm thick bottom layer of 1500 kg/m3 light aggregate concrete of tensile
strength fct20 = 2.7 MPa and conductivity λ = 0.6 W/mK
It is reinforced with 8 Y 8 mm slack deformed bars of tensile strength 550 MPa
with a cover thickness d = 15 mm. The anchorage length is 70 mm as for the massive
slab in Sect. 4.4.1.
The weight of this sandwich slab is about 205 kg/m2 .
At a fire test, the slab resisted a service load of 2.4 kN/m2 for 79 min standard fire
exposure, where it was calculated to resist it for 61 min [11].
The very light structure and low dead load allows long span, which are beneficial
for many applications. However, it may cause a problem, if a sound insulation of
55 dB or more is required.
The application of more densities of concrete at the same cross-section may have a
positive influence on the sound insulation. However, the sound insulation requirement
4.4 Light Concrete Slabs 65

may still lead to quite thick slabs even when you try to document it by testing a full
floor structure.

4.5 Super-Light Concrete Beams

4.5.1 Rectangular Super-Light Beams

The first super-light structures, which we considered after inventing the principle,
were ordinary rectangular beams.
When you design an ordinary concrete beam, you place the concrete and the
reinforcement as usual, and then you make a structural analysis to give an idea
about, how the forces approximately are distributed. Based on that, you prove that
you have sufficient material in the right places of the structure to resist the forces.
As described in Sect. 3.2, the super-light principle allows you to apply direct
engineering, where you at first decide an optimal distribution of forces and place
a strong material such as a strong concrete or reinforcement according to that as
shown in Fig. 3.4. You then fill the rest of the structure with a light concrete in order
to obtain the wanted outer shape for example with a horizontal upper side of a beam,
to get a fire protection of the strong materials, or perhaps a shear connection between
compression- and tension zone.
As explained in Sect. 3.1.2, a curved shape of the strong material would give rise
to an expensive and complicated curved mould according to the traditional way of
making structures. We saw that for example when we produced the double beam in
Fig. 7.10.
A simple solution to that problem could be to hang a flexible pipe upside down,
cast the strong concrete in the pipe and cast it into the light concrete as shown in
Fig. 3.5.
You then obtain the beam in Fig. 3.4 by turning it upside down.
Another solution could be to apply a pearl-chain structure as explained in Chap. 6
creating a curved optimized compression zone of strong concrete from a number
of straight mass-produced elements. This curved structure can then be cast into a
light concrete with the outer shape of the beam, which for example could be double
cantilevered as shown in Fig. 7.10.
When the cross-section becomes high and slender, the beam becomes a hung-up
wall, and you may apply the super-light principles and pearl-chain elements to that
for example as shown in Fig. 6.6 and investigated by Lind [12]. If such high cross-
sections should function as beams and not have a wall function separating rooms, you
may obviously save material by providing the structure with holes, where strength
is not needed as for example shown in Fig. 7.20.
High and slender beam cross-sections may also be applied as ribs in prefabricated
girders for example with a double-T (TT) cross-section or cross-sections with shape
of the Greek letter pi (
).
66 4 Slabs and Beams

Fig. 4.24 Super-light beam system for halls

4.5.2 Super-Light Beam System for Halls

Figure 4.24 shows a principle of a system of prefabricated super-light beams for

large halls. The beams are cantilevered, and two of them are connected to a double
cantilevered beam for example with a length of 60 m. We tested double cantilevered
beams already in 2007 [13] before the super-light principle was published and before
pearl-chains were invented (Fig. 7.10). Later we applied double-cantilevered vaults
for bridge engineering as shown in Figs. 7.9 and 7.17.
The super-light cantilevered beam system was at first sketched in [14]. A V-shaped
cross-section for example made of a light concrete with density 1200 kg/m3 has a
curved bottom that serves as mould for casting a high-strength concrete. The curved
high-strength concrete serves as a compression zone. At the top of the cross-section,
we include two tension zones, which can be made for example as pretensioned zones
of ordinary concrete that can resist tension as unloaded compression with the large
stiffness of the concrete.
The tension zones of two half beams are bolted together, and the double
cantilevered beam is placed on top of a column, where a bearing joint receives
the inclined forces from the compression zones of the two half beams.
At the top of the double cantilevered beam, a number of secondary double
cantilevered beams made by the same principle can be placed transversely, as shown
to the right in Fig. 4.24. The figure also shows a proposal for a glazing vault between
the secondary beams.
The curved bottom of the V-shaped secondary beams may serve to lead rainwater
to the main beam that again may lead it to the columns, so that the roof can be drained
efficiently.
References 67

References

1. Castberg NA (2013) Architectural engineering to super-light structures–design, implementa-

tion and constructability. PhD Thesis, BIG Architects and Department of Civil Engineering,
Technical University of Denmark, 213p
2. Tassello A (2011) Load-bearing capacity of super-light slabs. MSc Project, Department of Civil
Engineering, Technical University of Denmark, 72p
3. Hertz KD, Castberg Christensen AJ (2014) Super-light concrete decks for building floor slabs.
Struct Concr J FIB 15(4):522–529. Ernst & Sohn
4. Halldórsson EE (2012) Load bearing capacity of super-light decks. MSc Thesis, Faculty of
Civil and Environmental Engineering, University of Iceland, 60p
5. Rocca N (2010) Risposta al fuoco di elementi strutturali superleggeri. MSc Thesis, Università
degli Studi di Roma La Sapienza, 209p
6. Hertz KD (2019) Design of fire-resistant concrete structures. ICE Publishing, Thomas Telford
Ltd. ISBN: 9780727764447. London, 254p
7. Christensen JE (2013) Acoustic design of super-light structures. PhD Thesis, Gronmij Ltd and
Department of Civil Engineering, Technical University of Denmark. 140p
8. ISO (International Organization for Standardization) (1975) ISO 834 Fire-resistance tests.
Elements of building construction. ISO Geneva, Switzerland, 25p
9. Persson K, Raaschou L (2017) Super-light multi-story car park. MSc Project, Department of
Civil Engineering, Technical University of Denmark, 230p
10. Halding PS, Hertz KD (2020) Design for disassembly of super-light structures. RILEM spring
convention 2020
11. Hertz KD (2003) Documentation for calculations of standard fire resistance of slabs and walls
of concrete with expanded clay aggregate. Report R-048. Department of Civil Engineering,
Technical University of Denmark. ISBN 87–7877–108–0. Lyngby, 43p
12. Lind FL (2018) Super-light wall elements. MSc Project, Department of Civil Engineering,
Technical University of Denmark, 82p
13. Larsen F (2007) Light concrete structures. MSc Project, Department of Civil Engineering,
Technical University of Denmark, 117p
14. Hertz KD (2009) Super-light concrete with pearl-chains. Mag Concr Res 61(8):655–663.
Thomas Telford Ltd.
Chapter 5
Columns and Walls

Abstract The chapter provides an in depth method with examples on how to calcu-
late the capacity of columns or walls using layers of different types of concrete.
The force in a column can be guided via a strong concrete, while a lighter concrete
stabilizes the strong against buckling. Columns with Entasis are used as well.

5.1 Light Concrete Columns and Walls with Uniform

Cross-Section

5.1.1 Centrally Loaded Light Concrete Columns and Walls

Light concrete columns and walls can be made from light concrete with or without
steel reinforcement, but may also be composite and consist of parts of strong and
light concrete.
The composite cross-section may consist of several areas of strong and light
concrete. Figure 5.1 shows a simple example with a strong concrete section in the
centre. By placing a light concrete around the strong part, the designer may apply
the light concrete for stabilizing and protecting the strong concrete.
You will find this especially relevant if you apply a high-strength concrete suscep-
tible to explosive spalling in fire [1], where the light concrete keeps the temperature
below the required maximum of 350 °C. Furthermore, the light concrete avoids
buckling the small section that will be a result of application of a high strength
concrete.
In this Sects. 5.1 and 5.2, we deal with columns, where the cross-section is
uniform along the length of the column. This has been the usual way of designing
columns since the re-introduction of reinforced concrete in modern time, because it is
made with simple moulds, which were easy to construct and therefore economically
optimal.
However, it is not the most optimal solution statically to apply columns with
uniform cross-sections. Varying cross-sections may give better load-bearing capac-
ities, and with a new light concrete technology, where flexural textile moulds make

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 69

K. D. Hertz and P. Halding., Sustainable Light Concrete Structures, Springer Tracts
in Civil Engineering, https://doi.org/10.1007/978-3-030-80500-5_5
70 5 Columns and Walls

Fig. 5.1 Centrally loaded

light composite concrete
column

it affordable, such structures will often be economical and sustainable. We therefore

deal with these structures in Sect. 5.3.
For a plain column of one material, Euler (Fig. 5.2) presumed an elastic sinusoidal
deflection when he made his well-known formula for the load-bearing capacity of a
centrally loaded column.
π  dy π π 
y = δ sin x ⇒ = δ cos x ⇒
L dx L L

d2 y π2 π 
= κ = − δ sin x ,
dx2 L2 L

Moment equilibrium (inner moment = outer moment) of the cross-section at the

middle, where x = L/2, gives for an applied force F, and the deflection at the middle
δ

Fig. 5.2 William John Macquorn Rankine, Wilhelm Ritter, and Leonhard Euler. (Drawing by KD
Hertz)
5.1 Light Concrete Columns and Walls with Uniform Cross-Section 71

π2
δF = −κ E0 I ⇔ δF = δE0 I ⇔
L2
π2 E0 I
FE =
L2

In this derivation y is the elastic deflection in the length x.

κ is the curvature, L is the column length, and I is the moment of inertia of the
cross-section.
We have here presumed that the material is linear elastic.
From the derivation, you see that we not only obtain equilibrium between outer
and inner moment at the middle, but at any level x of the column.
 π  π2 E I  π2 π 
0
yFE = −κ(x)E0 I ⇔ δ sin x = δ sin x E0 I
L L2 L2 L

Euler’s expression represents equilibrium of the moment from the outer force
times the deflection and the inner elastic moment given as a product of flexural
stiffness (EI) and curvature.
As such, it only represents a requirement of stability and not a requirement of
sufficient compressive strength, which a designer must also consider, especially when
the slenderness of the column decreases.
As seen from the formula, the Euler force varies towards infinity, when the column
length L varies towards none. This is obviously wrong. Rankine (Fig. 5.2) therefore
suggested applying a variation limited to the compressive strength of the material.
Ritter (Fig. 5.2) then explained this by describing equilibrium of a deflected column
with a certain load for a material that has a curved lined stress–strain curve as seen
in Fig. 5.3 [2].

Fig. 5.3 Stress–strain curve

according to Ritter
72 5 Columns and Walls

We will obtain equilibrium for that if we apply the related reduced elastic modulus
equal to the inclination of the tangent to the stress–strain curve at that particular load
level.
We find this E-modulus from Ritter’s assumption
     
σ F dσ σ
E = E0 1 − = E0 1 − or = E0 1 −
fu Fu dε fu

In this expression, Fu represents the ultimate load in compression for the cross-
section and F the actual load on the column, which we want to find.
Solving the differential equation above, Ritter’s stress–strain curve becomes
   
E
− 0 ε
σ(ε) = fu 1 − e fu

We call the force at equilibrium in the deflected column for the Rankine force FR .
   
π2 IE0 F FR
FR = 1 − ⇔ FR = FE 1 −
L2 Fu Fu

From that, we get the simple expression

1 1 1
= +
FR Fu FE

FR is the ultimate load for the centrally loaded column.

This is a convenient expression for the so-called Rankine formula applied in
most textbooks and codes on concrete structures, although it usually appears as
expressions, which are hard to recognize.
Still, we have only assumed an ultimate strength, a Ritter stress–strain curve, and
a sinusoidal deflection as a basis for our structural calculations. We have derived
everything else.
In general, all parts of a column (strong concrete, light concrete, and reinforce-
ment) follow the same deflection, and all parts including modern reinforcement have
a curved stress–strain curve.
Hence, it is possible for you to find a Rankine load-bearing capacity for a column
consisting of reinforcing bars alone, strong concrete alone or light concrete alone.
Doing so, you presume that something keeps the mutual distance between the
different parts when they deflect. This means that you may add the contributions of
the cross-sectional areas of each material each weighted by its elastic modulus.
Ritter also added steel and concrete areas by multiplication of the contribution of
the steel area with a factor 10 that was an approximation for Es0 /Ec0 and fsu /fcu . We
apply the actual values of E0 and fu instead, when we add contributions of areas of
different materials.
5.1 Light Concrete Columns and Walls with Uniform Cross-Section 73

From that, you may get an extended Rankine formula for a centrally loaded
reinforced composite concrete column combining the three expressions.
This approach is similar to the expressions developed in [1] for calculating the
load-bearing capacity of fire exposed reinforced concrete columns. In these, the
impact of fire has changed the cross-sections so that they contain a variety of concrete
qualities and a variety of reinforcement qualities dependent on how much the fire
has heated each zone of a cross-section.
Here, s means steel, c strong concrete and lc light concrete.

1 1 1
= +
FR Fcu + Flcu + Fsu FcE + FlcE + FsE

This expression shows the load-bearing capacity FR of a centrally loaded

reinforced light composite concrete column.

5.1.2 Eccentrically Loaded Uncracked Columns and Walls

In the previous clause, we derived the Rankine formula for the load-bearing capacity
of a centrally loaded reinforced composite concrete column.
However, we apply eccentric load to most columns as illustrated in Fig. 5.4.
We will therefore check the stresses in a deflected, eccentrically loaded column
to see whether it is stable for a given load F, and eccentricity e.
Here, we apply the following expressions for the total axial- (EA) and flexural
stiffness (EI) of a composite cross-section at a certain load level related to a common
axis of bending

Fig. 5.4 Eccentrically

loaded column
74 5 Columns and Walls

EA = Ec Ac + Elc Alc + Es As and EI = Ec Ic + Elc Ilc + Es Is

For an unloaded cross-section the same properties become

EA0 = Ec0 Ac + Elc0 Alc + Es0 As and EI0 = Ec0 Ic + Elc0 Ilc + Es0 Is

The ultimate compressive strength of the total cross-section is also needed in the
formulas.
For an ultimate strain of concrete of more than 0.35% and the steel yielding above
0.2% + fsu /Esu = approximately 0.4%, you may assume the strain to be the same
and

Fu = Ac fcu + Alc flcu + As fsu

We apply the tangent elastic module of the curved working curve suggested by
[2] for each of the concrete qualities in the cross-section. We find the tangent elastic
modulus as inclination of the tangent to the curved working curve for the strain of
the entire cross-section subjected to the axial central load F.
We can then find a stress-distribution for the eccentricity e using the tangent
elastic modulus of each area (Fig. 5.5) and thereby we assume that all parts of the
cross-section deform according to their position and that plane cross-sections remain
plane.
We make the reasonable presumption that the entire cross-section will fail when
the strongest part fail. We therefore consider each of the different concrete qualities
to follow a Ritter stress–strain curve with the same ultimate strain. This means that
we consider the ratio of the reduction of the elastic moduli to be the same for all
types of concrete, when we apply a load F, on the column. In case the concretes have

Fig. 5.5 Stress–strain

curves for different concretes
in the same cross-section
5.1 Light Concrete Columns and Walls with Uniform Cross-Section 75

different ultimate strains, we apply the smallest, which will also be the one of the
strongest concrete.
We can therefore find the reduced elastic modulus for each of the different
materials as
   
σ F
E = E0 1 − = E0 1 −
fu Fu

This means that we can apply this reduction to the entire axial and flexural stiffness.

EA = Ec Ac + Elc Alc + EAs ⇔

 
F
EA = (Ec0 Ac + Elc0 Alc + Es0 As ) 1 − ⇔
Fu
 
F
EA = EA0 1 −
Fu

And similarly
 
F
EI = EI0 1 −
Fu

The Euler strength of the total composite cross-section then becomes

π2 EI0
FE =
L2
To assess the curvature and deflection δ for the eccentric loaded column, we apply
the same sinusoidal deflection curve as Euler because we, just like Euler, claim that
the difference of the result between using that and the actual curve will be modest.
We then apply a condition of equilibrium similar to the one Euler used for a central
loaded column but now with a lever arm, e + δ, to the force, F, in a deflected condition
at the midpoint (Fig. 5.4). This means that we require the external moment of the
force with lever arm e + δ to be equal to the internal moment of the cross-section
subjected to the curvature for the deflection δ.
 
π2 π2 EI0 F
(e + δ)F = 2 δ EI ⇔ (e + δ)F = 1− δ⇔
L L2 Fu
   
F F F F F
e = 1− − δ⇔e = 1− δ⇔
FE Fu FE FE FR
F
FE
1− F
FR
+ F
FE
1− F
Fu
δ=e ⇔ (e + δ) ⇔ e δ(e + δ) ⇔ e
1− F
FR
1− F
FR
1− F
FR
76 5 Columns and Walls

This expression calculates the lever arm e + δ in a deflected condition, of a

column loaded by a force F with an eccentricity e. The eccentricity e is increased by
a sinusoidal deflection of magnitude δ in the midpoint in equilibrium with the force.
We can then check the load-bearing capacity of a composite column for a given force
F by checking that the total cross-section can withstand the moment (e + δ) F and
the normal force F. If that is the case, the column can carry the normal force F with
an initial eccentricity e with respect to instability.

F
1−
Fu
(e + δ) = e
F
1−
FR

If we know the normal force F and eccentricity e, we have to design the cross-
section of the column to resist F and moment (e + δ) F. In case we want to calculate
the load-bearing capacity of a given column with a given eccentricity, we have to
guess the force F and check the cross-section for that, and within a couple of iterations
find the critical F.
In both cases, we can apply different methods for calculating the composite cross-
section just as we can for beams with a normal force.
(1) We can find an elastic stress-distribution and compare the stresses at the edges
with the strength of the materials in compression and tension.
This is usually on the safe side compared to a plastic approach.
(2) We can find a plastic stress-distribution presuming that no tensile stresses are
applied.
(3) We can find a plastic stress-distribution with application of tensile stresses.
This is applicable for beam-columns, where the bending moment is large.
In the first case, we calculate stresses at the edges of each part of the cross-section
such as strong concrete, light concrete and reinforcement, for the column subjected
to a certain load F and moment (e + δ) F and compare the stresses with the strength.
As an example, for the strong concrete we compare the compressive stress at the
distance ycc from the centroid of the composite cross-section with the compressive
strength fcc

F Ec F(e + δ)Ec
σc1 ≤ fcc ⇔ + y ≤ fcc ⇔
 EA
 EI  cc 
F Ec0 1 − FFu F(e + δ)Ec0 1 − FFu
  +   ycc ≤ fcc ⇔
EA0 1 − FFu EI0 1 − FFu

FEc0 F(e + δ)Ec0

+ ycc ≤ fcc
EA0 EI0
5.1 Light Concrete Columns and Walls with Uniform Cross-Section 77

Similarly, we compare the tensile stress at the distance yct with the tensile strength
fct

F Ec F(e + δ)Ec
σc2 ≥ −fct ⇔ − y ≥ −fct ⇔
  EA EI  ct 
F Ec0 1 − FFu F(e + δ)Ec0 1 − FFu
  −   yct ≥ −fct ⇔
E0 A 1 − FFu E0 I 1 − FFu

F Ec0 F(e + δ)Ec0

− yct ≥ −fct
EA0 EI0

We then do the same check for light concrete and reinforcement.

5.1.3 Columns and Walls with Plastic Compression

In this case, we consider a column or wall with height L and a massive cross-section
of for example a light concrete. We presume that the cross-section should be able to
resist a force with eccentricity e in a deflected condition by a plastic distribution of
ultimate compressive stresses symmetrically placed around the eccentric force.
We denote the height of the concrete cross-section h in the direction of the
eccentricity e. The width is always c no matter if the height is larger or not.
In order to estimate the deflection we calculate the compressive strength Fu and
the flexural stiffness EI0 using the initial elastic modulus E0 as shown in Sect. 5.1.1.
From that, we find the Euler force

π 2 EI0
FE =
L2
The result will be a Rankine force FRc from the formula

1 1 1
= +
FR Fu FE

The total lever arm at the middle of the column is then the sum of the eccentricity
e and the deflection δ. In Sect. 5.1.2 we found an expression for (e + δ) as

1− F
Fu
(e + δ) = e
1− F
FR

If we want to find the critical load F for a column or wall as shown in Fig. 5.6,
we can presume a value of F in order to find the deflection δ. The calculation is an
iteration.
78 5 Columns and Walls

Fig. 5.6 Plastic compression

on a massive column or wall

However, when we apply the expression for the lever arm (e + δ) for calculation of
the compressive strength symmetrically around the eccentric force F we can obtain
an explicit expression of the critical load. This means that we can avoid iteration

F h − 2(e + δ) F e1 − F
Fu
= ⇔ =1−2 ⇔
Fu h Fu h 1− F
FR
F e
1− =2 ⇔
FR h
 e
F = FR 1 − 2
h

5.1.4 Beam-Columns and Cracked Walls

We call the structure a beam-column, if we apply a moment load so large that the
tension zone of the concrete cracks and we mainly rely on resisting tension in the
reinforcement.
This means that the column cannot carry its load as an uncracked structure
according to Sects. 5.1.2 and 5.1.3.
For uncracked columns, we considered the stiffness of the entire cross-section
calculating the deflection and we found the ultimate moment capacity of the part of
the cross-section with strong concrete and reinforcement. This is in accordance with
5.1 Light Concrete Columns and Walls with Uniform Cross-Section 79

the calculations we make for the ultimate load-bearing capacity of composite beams
and slabs and for uncracked composite columns.
For cracked beam-columns, we also find the ultimate moment resistance of the
strong concrete alone and compare this with the moment in the column at the ultimate
deflection, where the strains in the cross-section correspond to the ultimate moment.
The column load F has a depth dF from the compressed edge of the strong concrete.
If the column force has an eccentricity e from a centroid in a depth ycc , this is

dF = ycc − e

However, all you need to know is the depth dF of the column load before deflection.
The depth dF is positive if the column load F acts within the strong concrete before
the column deflects and negative, if the moment is so large that it acts beyond the
compressed edge.
We call the width of the strong concrete across the direction of the depth dF for
cc .
cc does not need to be the smallest dimension of the strong concrete, but often it
is.
The strong concrete may also consist of more sections and then cc is the width
of the section with the compressed edge, where the compressive stresses will be in
a cracked column.
For a beam-column with tension in the tensile reinforcement, we can assess the
ultimate deflection δu of a reinforced cross-section from the curvature calculated for
the ultimate resistance of the cross-section at the critical column load F.
We apply the ultimate strain for normal concrete εcu = 0.0035 at the compressed
edge of the strong concrete, and the yield strain εsy and the yield force of a tension
reinforcement Fs2 in the depth ds2 from the compressed edge of the strong concrete
(Fig. 5.7).

Fig. 5.7 Beam-column

cross-section with eccentric
load F before deflection
80 5 Columns and Walls

fs
εsy = 0.002 +
Es0

Still presuming a sinusoidal deflection, we find the ultimate value δu of this

deflection as
εcu + εsy 2
δu = L
π 2 ds2

The deflection δu reduces the depth of the column load from the compressed edge,
which means that the deflection increases the moment of the column load. In some
cases, the column force F may act beyond the compressed edge and its depth dF is
negative. In such cases, the deflection still reduces the lever arm and increases its
negative value.
For the ultimate deflection, the external moment load of the critical column load F
calculated about the compressed edge of the strong part of the cross-section becomes

M = −F(dF − δu )

Reinforcement will yield in compression before the concrete around it fails. There-
fore, we can also obtain a yield force Fs1 in the compression reinforcement at depth
ds1 near the compressed edge (Fig. 5.8).

Fig. 5.8 Cracked composite

column with ultimate
deflection δu
5.1 Light Concrete Columns and Walls with Uniform Cross-Section 81

The total force in the reinforcement is

Fs = Fs2 − Fs1 (Positive in tension)

The moment contribution about the compressed edge of Fs2 in tension and Fs1 in
compression is then

Ms = ds2 Fs2 − ds1 Fs1

You then find the moment contribution about the compressed edge of the force Fc
= F + Fs in the strong concrete section of width cc and compressive strength fcc

(F + Fs )
Mc = −(F + Fs )
2cc fcc

Inserting these expressions in the condition that the external moment of the load
should be equal to the internal moment in the deflected cross-section, you find a
second-order equation in the critical column load F

M = Mc + Ms ⇔ −M = −Mc − Ms ⇔
 
(dF − δu )F2cc fcc = F2 + 2FFs + F2s − Ms 2cc fcc ⇔
F2 + (2Fs − (dF − δu )2cc fcc )F + F2s − 2cc fcc Ms = 0

In order to solve this, we define a parameter b and find the critical load F as

b = 2Fs − (dF − δu )2cc fcc


−b + b2 − 4F2s + 8cc fcc Ms
F=
2

This is an expression of the critical load F of an eccentrically loaded reinforced

concrete column with tensile stresses in the reinforcement in a deflected condition
– a beam-column.
You can then compare the result of F with the applied load Fload and the eccentricity
measured as the depth dF of the load from the compressed edge of the strong concrete
before deflection.

5.1.5 Design of Columns and Walls—Summary

When we design an eccentric loaded column or wall, we can chose how we want the
column to react to the maximum column force in the deflected condition.
If we want that the column or wall to be safe form cracking, we apply method (1).
82 5 Columns and Walls

If a massive column or wall should carry the load in compression, we apply method
(2).
If we allow cracks and ultimate stresses in strong concrete and reinforcement, we
apply method (3).
Method (1). Uncracked according to Sect. 5.1.2. Here we apply a column force F
that we either know or guess with an eccentricity e.
We find the ultimate compressive load and the flexural stiffness

Fu = Ac fcu + Alc flcu + As fsu and EI0 = Ec0 Ic + Elc0 Ilc + Es0 Is

Using the ultimate compressive strength fu and the initial modulus of elasticity
E0 for each material, we find the Euler force FE and from that, the Rankine force FR

π2 EI0 1 1 1
FE = and = +
L2 FR Fu FE

The Euler force FE is the critical load of the column or wall, if the eccentricity is
e = 0 mm and the materials are linear elastic. The Rankine force FR is the critical
load if we use the tangent elastic moduli of the stress–strain curves at the load level
considered.
From that, we find the lever arm equal to the initial eccentricity plus the deflection
e + δ:

F
1−
Fu
(e + δ) = e
F
1−
FR

We then check that the stresses do not exceed the strength anywhere in the deflected
composite cross-section as for example the maximum tensile stress of the strong
concrete at the edge in tension at the distance yct from the centroid of the composite
cross-section
FEc0 F(e + δ)Ec0
− yct ≥ −fct
EA0 EI0

We apply the initial elastic modulus for each material E0 to find the initial axial and
flexural stiffness EA0 and EI0 . We do that, because we have proved that the tangential
elastic modulus and the axial and flexural stiffness of the loaded cross-section would
give the same result.
This is because the stiffness of the different materials relate equally to each other
in a loaded and in an unloaded cross-section.
Method (2). In case we consider a massive column or wall, we can avoid guessing
a critical load F, if we consider a plastic uncracked cross-section according to
Sect. 5.1.4.
5.1 Light Concrete Columns and Walls with Uniform Cross-Section 83

The capacity is then:

 e
F = FR 1 − 2
h
Method (3). A cracked cross-section according to Sect. 5.1.4.
Here, we find the ultimate plastic moment of the strong concrete and the rein-
forcement, as we would do for a beam. We apply a deflection δu corresponding to
the ultimate plastic moment:

εcu + εsy 2
δu = L
π 2 ds2

Thereby we find the ultimate critical load of the column in the cracked condition,
where we consider it as a beam with normal load F in a depth before deflection dF
from the compressed edge of the strong concrete and with a width cc . We calculate
dF as positive for F in the cross-section and negative for larger moments, where F is
placed beyond the compressed edge.

b = 2Fs − (dF − δu )2cc fcc


−b + b2 − 4F2s + 8cc fcc Ms
F=
2
Here, we do not apply the strength and stiffness of the light concrete (as we also
do not for plastic moments for beams. The result is therefore on the safe side for the
entire column.

5.2 Examples Columns and Walls with Uniform

Cross-Section

5.2.1 Example of a Tested Column

As an example, we consider a L = 2.0 m long composite concrete column (Fig. 5.9)

with a central core of strong concrete hc is 50 mm and cc is 90 mm. The ultimate
compressive strength fcc is 45 MPa, the tensile strength fct is 3.8 MPa, and the initial
elastic modulus Ec0 is 36 GPa. The column is hinged in both ends, and it has been
full-scale tested by [3].
Around the central strong core is cast a light aggregate concrete of density
950 kg/m3 with cross-sectional properties: hlc = 180 mm, clc = 190 mm, ultimate
strength flcc = 5.0 MPa, tensile strength flct = 0.5 MPa and initial elastic modulus
Elc0 = 5.0 GPa (Fig. 5.10).
84 5 Columns and Walls

Fig. 5.9 Composite column test

Fig. 5.10 Cross-section and cracked composite column

Six Y10 reinforcing bars with Ds = 10 mm, yield strength fs = 550 MPa, and
elastic modulus Es0 = 200 GPa are placed in the light aggregate concrete with cover
thickness cs = 25 mm.
Their area As , distance from centre ys , and moment of inertia Is are

π hlc
As = 6 D2s , ys = − cs , Is = As y2s
4 2
A load is applied with an eccentricity e = 10 mm. We calculate the load-bearing
capacity and compare it to the capacity found in the laboratory test.
5.2 Examples Columns and Walls with Uniform Cross-Section 85

The area and moment of inertia of both the strong concrete and light aggregate
concrete and the steel is
1
Ac = hc cc , Ic = cc h3
12 c
1
Alc = hlc clc − Ac , Ilc = clc h3lc − Ic
12
The ultimate strength of the strong- and light concrete and the steel is

Fcc = Ac fcc = 203 kN, Flcc = Alc flcc = 149 kN, Fs = As fs = 259 kN

The total ultimate strength is

Fu = Fcc + Flcc + Fs = 610 kN

The axial stiffness of the column is

EA0 = Ec0 Ac + Elc0 Alc + Es0 As = 405 MN

The flexural stiffness is

EI0 = Ec0 Ic + Elc0 Ilc + Es0 Is = 889 kN m2

The Euler strength of the strong- and light concrete and steel is

π 2 EIc0 π 2 EIlc0 π 2 EIs0

FcE = = 83 kN, F lcE = = 1128 kN, F sE = = 983 kN
L2 L2 L2
The total Euler strength is FE = FcE + FlcE + Fs = 2193 kN.
The Rankine force of the column becomes
1
FR = = 477 kN
1
Fu
+ 1
FE

If we guess a column force F = 266 kN, we can calculate the corresponding

deflection, where the external moment from the axial load with eccentricity and
deflection is in equilibrium with the internal moment with an elastic distribution of
stresses at the composite cross-section.

1− F
Fu
(e + δ) = e = 12.74 mm
1+ 1
FR

With this balancing moment, we can find the maximum and minimum stresses at
the edges of the strong and the light concrete and in the reinforcing bars.
86 5 Columns and Walls

FEc0 F(e + δ)Ec0 hc √

σc1 = + = 27.1 MPa ≤ fcc = 45 MPa
EA0 EI0 2
FEc0 F(e + δ)Ec0 hc √
σc2 = − = 20.2 MPa ≥ −fct = −3.8 MPa
EA0 EI0 2
FElc0 F(e + δ)Elc0 hlc √
σlc1 = + = 5.00 MPa ≤ flcc = 5 MPa
EA0 EI0 2
FElc0 F(e + δ)Elc0 hlc √
σlc2 = − = 1.57 MPa ≥ −flct = −0.5MPa
EA0 EI0 2
FEs0 F(e + δ)Es0 √
σs1 = + ys = 181.0 MPa ≤ fs = 550MPa
EA0 EI0
FEs0 F(e + δ)Es0 √
σs2 = − ys = 81.9 MPa ≥ −fs = −550MPa
EA0 EI0

This means that the column can carry at least 266 kN with the eccentricity 10 mm,
if the force acts on the total composite cross-section and we consider an elastic stress
distribution according to Sect. 5.1.2.
If we instead only consider the load-bearing capacity according to Sect. 5.1.4
for a cracked cross-section with ultimate moment in the reinforcement and strong
concrete. Then we get δu = 37.2 mm, Fs = 0 kN, Ms = 16.9 kNm

fs
εsy = 0.2% + = 0.00475; εcu = 0.00350; ds2 = 0.09 m;
Es0
εcu + εsy 2 hc
δu = L = 0.0372 m; dF = − e = 0.015 m
π 2 ds2 2
Ms = ds2 Fs2 − ds1 Fs1 = (0.09 m − (−0.04 m))Fs2 = 16.9 kNm

and

b = 2Fs − (dF − δu )2c fcc = 179 kN


−b + b2 − 4F2s + 8c fcc Ms
F= = 290 kN
2
We can therefore expect the load-bearing capacity of the column to be 266 kN
before it cracks and 290 kN in an ultimate cracked condition.
However, the tested column was constructed so that the force acted on a part of
the central strong concrete section sticking out of the column as seen in Fig. 5.10.
We therefore have to find the minimum load-bearing capacity as a minimum of
the critical load of the column and the strength of this part subjected to the eccentric
load without deflection.
5.2 Examples Columns and Walls with Uniform Cross-Section 87

Fig. 5.11 Failed strong column

The capacity in pure eccentric compression of the projected part will be

 
2e
Fcc 1 − = 121.5 kN
hc

This capacity is less than the critical load of the column and therefore determines
the capacity of the whole structure at the test.
We made two tests with such column [3], and saw failures at 122 kN and 128 kN.
The failures occurred at the projecting strong concrete in both tests, and no damage
was observed on the rest of the columns (Fig. 5.10).
Then, we tested a column consisting of the central, unreinforced, strong concrete
part alone without the enclosing, stabilizing light aggregate concrete (Fig. 5.11). We
still applied a load with eccentricity e = 10 mm. The Rankine force of this column
is:
1
FcR = = 59 kN
1 1
+
Fcc FcE

The deflection, where the internal moment is in equilibrium with the external is

1− F
Fcc
(e + δ) = e = 14.53 mm
1− 1
FcR

For a safe-side calculation based on maximum and minimum stresses of an elastic

distribution we can guess a capacity of F = 24 kN, and then we get
88 5 Columns and Walls

F F(e + δ) hc
σc1 = + = 14.02 MPa ≤ fcc = 45 MPa
Ac Ic 2
F F(e + δ) hc
σc2 = − = −3.8 MPa ≥ −fct = −3.8 MPa
Ac Ic 2

If we instead find the ultimate load-bearing capacity of the column based on a

plastic stress-distribution in the deflected condition as shown in Sect. 5.1.3, we get:
 
2e
FcR 1 − = 35.4 kN
hc

We should therefore expect that the tested column would have a failure load
between 24 kN and 35.4 kN. In the test, the column failed at 29.5 kN (Fig. 5.11).
As seen from the example, the composite column, where a light concrete stabilizes
the strong, had a load-bearing capacity that was four times larger than the same
column consisting of only the strong concrete. Furthermore, if we had not applied
the load on a weak projecting part of the strong concrete, we could expect it to have
been nine times stronger.
We will now consider the composite column loaded with varying eccentricity and
without failure at the projecting strong part. With an increasing eccentricity from 0 to
60 mm, we find the corresponding load-bearing capacity considering the uncracked-
or the cracked condition.
Figure 5.12 shows the result. When the eccentricity increases to 25 mm, the critical
condition for the uncracked column changes from crushing of the compression zone
of the light concrete to cracking of the tension zone of the light concrete. You can
see that as a change in inclination of the curve at this point.

Fig. 5.12 Theoretical capacity of test column with variable eccentricity

5.2 Examples Columns and Walls with Uniform Cross-Section 89

For eccentricities less than 5 mm, the column will fail as soon as it reaches the
limit for the uncracked condition. For larger eccentricities, the column can deflect
further and carry a larger force than the limit for the uncracked condition.

5.2.2 Super-Light Wall Design

Lind [4] investigated a number of designs of super-light wall elements comprising

sandwich walls as described in Sect. 5.2.3 and walls, where strong concrete
constituted frames with holes filled out by light concrete.
The main problem for sustainable light wall elements is usually a requirement of
sound insulation for separating walls. The Danish Building Regulations requires for
example 55 dB, and this means that a massive wall should have a mass of 440 kg/m2 .
Super-light sandwich walls may fulfill the requirement with a mass of only 350 kg/m2
because the concrete materials of different density oscillates with different eigen-
frequencies, transforming some of the sound energy to heat, as we can see from the
sound tests on super-light deck elements described in Sect. 4.1.4.
For this reason, sandwich walls or sandwich like walls may represent more
promising design solutions for domestic buildings than walls designed as frames
of strong concrete with light fillings. However, they can be interesting, where sound
requirements do not represent a design limit.

5.2.3 Example Sandwich Wall

We now consider a part c = 1 m of a L = 3.0 m high sandwich wall (Fig. 5.13)

consisting of two layers of strong concrete (each tc = 30 mm thick) separated by a
layer of light concrete of thickness hlc = 90 mm.
The wall thickness is then hc = 2·30 + 90 = 150 mm.

Fig. 5.13 Sandwich wall cross-section

90 5 Columns and Walls

The strong concrete has an ultimate compressive strength fcc = 55 MPa, a tensile
strength fct = 4.2 MPa, and an initial elastic modulus Ec0 = 38 GPa.
The light aggregate concrete is of density 700 kg/m3 and has an ultimate strength
flcc = 3.0 MPa, a tensile strength fct = 0.3 MPa, and an initial elastic modulus Ecl0
= 3.0 GPa.
An Y8 reinforcement (Ds = 8 mm) with the strength fs = 550 MPa is placed per
150 mm in each “flange” of strong concrete.
For these data input we get: Fcc = 3300 kN, Flcc = 270 kN and Fs = 369 kN.
In addition, the Euler force becomes

FcE = 9189 kN, FlcE = 200 kN and FsE = 529 kN. In total FE = 9918 kN.

It is observed that the two flanges of strong concrete are the main load-carrying
parts of the cross-section.
We find the ultimate compressive force and the Rankine force

Fu = 3939 kN, FR = 2819 kN.

For an eccentricity of e = 20 mm we guess an uncracked capacity of F = 1890

kN.
From that we find get (e + δ) = 31.6 mm and check the maximum stresses

σc1 = 45.6 MPa ≤ fcc = 55 MPa

σc2 = 7.96 MPa ≥ fct = −4.2 MPa
σlc1 = 3.00 MPa ≤ flcc = 3.0 MPa
σlc2 = 1.22 MPa ≥ flct = −0.3 MPa
σs1 = 220.0 MPa ≤ fs = 550 MPa
σs2 = 61.7 MPa ≥ fs = −550 MPa

The guessed force is the limit for compressive failure of the light concrete.
Furthermore, we find the cracked deflection δu = 55.7 mm, Ms = 22.1 kNm and
b = 79.9 kN.
From that, we determine the cracked capacity of 1520 kN. This means that the
sandwich wall with eccentricity of 20 mm will fail when it is loaded to crushing,
because it does not get a larger critical load when the deflection increases and the
cross-section cracks.
If we vary the eccentricity e, we get the results shown in Fig. 5.14.
The curve for uncracked failure changes inclination at e = 40 mm, where the limit
changes from compressive failure of the light concrete to tensile failure of the strong
concrete.
For this wall the reinforcement will increase the load-bearing capacity for eccen-
tricities less than e = 10 mm and larger than e = 80 mm. In these intervals, the
critical load increases, when the deflection increases and the cross-section cracks.
In between these limits the wall will break, when the first crushing occurs.
5.2 Examples Columns and Walls with Uniform Cross-Section 91

Fig. 5.14 Theoretical capacity of sandwich wall with variable eccentricity

5.2.4 Example Wall and Column with Circular Light Cores

If the central light concrete filling of a sandwich wall is replaced with a series of
circular light concrete cores as shown in Fig. 5.15, you obtain the benefit of a sandwich
structure and in addition an improved axial stiffness. The profile may be beneficial
for columns with one or more circular cavities.
We now consider a part c = 1 m of a L = 3.0 m high wall with mainly the same
data as the sandwich wall in Example 5.2.3 tc = 30 mm (thickness flanges) and
circular light concrete parts with diametre Dlc = 90 mm. The total cross-sectional
height is still hc = 2·30 + 90 = 150 mm.
As for the sandwich wall we apply a strong concrete with fcc = 55 MPa, fct =
4.2 MPa and Ec0 = 38 GPa. The light concrete of density 700 kg/m3 has flcc =
3.0 MPa, flct = 0.3 MPa and Ecl0 = 3.0 GPa.
Y8 reinforcement Ds = 8 mm has the strength fs = 550 MPa and is placed per
120 mm (each segment is a 90 mm circle plus a 30 mm web equal to tc ).

Fig. 5.15 Wall and column with circular light cores

92 5 Columns and Walls

For these data we get: Fcc = 5334 kN, Flcc = 159 kN and Fs = 461 kN. The Euler
force is

FcE = 10,602 kN, FlcE = 88 kN and FsE = 662 kN. In total FE = 11,352 kN.

From that we determine the ultimate compression force and the Rankine force

Fu = 5954 kN, FR = 3906 kN.

The strong concrete is the main load-carrying part of the cross-section as it also
was for the sandwich wall.
For an eccentricity of e = 20 mm we guess an uncracked capacity of F = 2550
kN.
From that we find get (e + δ) = 32.9 mm and check the maximum stresses.

σc1 = 47.3 MPa ≤ fcc = 55 MPa

σc2 = 1.02 MPa ≥ fct = −4.2 MPa
σlc1 = 3.00 MPa ≤ flcc = 3.0 MPa
σlc2 = 0.81 MPa ≥ flct = −0.3 MPa
σs1 = 224.5 MPa ≤ fs = 550 MPa
σs2 = 29.7 MPa ≥ fs = −550 MPa

The guessed force is the limit for compression failure of the light concrete.
Furthermore, we find the cracked deflection δu = 55.7 mm, Ms = 27.7 kNm and
b = 79.9 kN.
From that, we find the cracked capacity of 1704 kN. This means that the wall with
eccentricity of 20 mm will fail in case it is loaded to crushing, because an increase
of deflection will not increase the critical load.
If we vary the eccentricity e, we get the results shown in Fig. 5.16.
When the eccentricity increases to more than e = 30 mm, tension failure of the
strong concrete happens before compression failure of the light concrete.
This wall has a larger load-bearing capacity at the ultimate deflection for all eccen-
tricities than the similar sandwich wall in Example 5.2.3. It has a larger cross-section
of strong concrete, and therefore it has a larger uncracked load-bearing capacity than
the sandwich wall for small eccentricities.
We now consider the column to the right in Fig. 5.15. It is quadratic with side
lengths of 200 mm and a central light concrete cavity of diameter 140 mm. It has four
Y8 reinforcing bars and length L = 3 m. We consider it to have the same materials
as applied in the wall. This leads to Fcc = 1353 kN, Flcc = 46 kN, Fs = 111 kN, FcE
= 4770 kN, FlcE = 62 kN and FsE = 319 kN. In total FE = 5151 kN.
From that we find: Fu = 1510 kN, FR = 1168 kN.
For an eccentricity of e = 20 mm we guess an uncracked capacity of F = 722 kN
and from that we get (e + δ) = 27.3 mm.
5.2 Examples Columns and Walls with Uniform Cross-Section 93

Fig. 5.16 Theoretical capacity of wall with circular light cores and variable eccentricity

From the eccentricity e = 50 mm the failure mode in the uncracked condition

changes from compression in the light concrete to tension in the strong concrete.
At all eccentricities, the cracked condition has a larger capacity, which means
that the column can be loaded beyond the limit, where the first cracking is observed
(Fig. 5.17).

Fig. 5.17 Capacity of column with circular light core and variable eccentricity
94 5 Columns and Walls

5.3 Light Concrete Columns and Walls with Entasis

5.3.1 Centrally Loaded Columns and Walls with Entasis

Entasis is according to Vitruvius [5] the Greek expression for a light convex curved
surface of a column, so that it is thicker at the middle than at the ends. This gives
an optical effect counteracting the impression that a straight column may appear
more slender at the middle, and many authors refer to this as a reason why ancient
architects applied it.
An engineering reason may be that a column with Entasis has the largest stiffness
and cross-sectional resistance at the middle, where the deflection and the curva-
ture in a deflected condition is the largest for a simply supported column. In other
words, application of Entasis counteracts deflection by placing material, where the
column requires it the most and thereby serve a purpose of reducing the total material
consume.
The Greek word Entasis means “tension” and is formed of “en- “ and “teinein”,
which means to stress. This is probably because a column with Entasis gives an
optical impression of tension or swelling. However, for a light concrete structure
this meaning may become physical. If we cast a concrete column with Entasis in a
mould, this “swelling” gives a tensile stress in the mould material and the convex
curved shape becomes natural for counteracting the mould pressure in tension.
Furthermore, application of a light concrete reduces the mould pressure and the
tensile stress for example to a quarter of that of an ordinary concrete. This means
that we can apply quite different mould materials such as light textiles and obtain
more elegant, economical, and sustainable constructions.
From a point of view of engineering optimization, we would like to utilize all
cross-sections of the column better than we do for an ordinary straight column,
where the cross-section at the middle is subjected to the largest curvature.
As explained in Sect. 5.1.1, you derive the Euler force based on a presumption that
the deflection of a simply supported column with a uniform cross-section follows
a sinus curve. This implies that the curvature of the column is zero at the ends and
maximum at the midpoint. It is a reasonable presumption since the moment in a
centrally loaded column is proportional to the deflection.
However, when you design your column with an entasis, you obtain a reduction
of the curvature at the midpoint compared to a column with uniform cross-section
because you have increased the flexural stiffness. Since most columns are calculated
for an eccentric load, it would be more reasonable to apply a presumed deflection
with a uniform curvature κ along the height of the columns with entasis so that all
parts of the column contribute equally much to resist curvature and deflection.
This presumption gives us a precondition for the deflection δ – y(x), where δ is
the deflection at the middle of the column, where we consider x = 0 m. The ends of
the column are then at x = L/2 and x = - L/2, where y(L/2) = δ. (Fig. 5.18)
5.3 Light Concrete Columns and Walls with Entasis 95

Fig. 5.18 Circular column

with Entasis

d2 y dy 1
2
=κ⇔ = κx + a ⇔ y(x) = κx2 + a x + b
dx dx 2

Symmetry gives a = 0 and the precondition that y(0) = 0 gives b = 0 and we get:

1
y(x) = κx2
2
   
L 1 L 2 δ
y = κ ⇔κ=8 2
2 2 2 L

At first, we consider the materials linear elastic with elastic flexural stiffness
E0 I(x). External equals internal moment gives
 
x2 δ
F(δ − y(x)) = E0 I(x) ⇔ F δ − 4 2 δ = 8 2 E0 I(x)
L L

For x = 0 you get the load-bearing capacity of a centrally loaded column or wall
with Entasis, linear elastic material and constant curvature

8E0 I(0)
F0 =
L2

This value is 81% of the similar Euler force, and as explained above, it is more
reasonable to apply for eccentric loaded columns with Entasis.
As for the Euler force in Sect. 5.1.1 for columns with uniform cross-section,
it is now possible to replace the linear elastic materials with materials with curved
working curves according to the Ritter assumption. We consider the flexural stiffness
at the middle of the column (where x = 0 m) and reduce the elastic modulus in that
96 5 Columns and Walls

for the load of the Rankine force FR0 of the column or wall with entasis. This gives
us
   
8IE0 F FR0
FR0 = 2 1 − ⇔ FR0 = F0 1 −
L Fu Fu

From that, we get the simple expression

1 1 1
= +
FR0 Fu F0

As for columns and walls with uniform cross-sections, we can consider the cross-
sections as composite structures with

Fu = Ac fcu + Alc flcu + As fsu and EI0 = Ec0 Ic + Elc0 Ilc + Es0 Is

Similar to columns and walls with uniform cross-section, we can derive an expres-
sion for the total deflection of a column or wall with an Entasis and a constant
curvature.
 
8 8E0 I(0) F
(e + δ)F = 2 δ EI(0) ⇔ (e + δ)F = 1− δ⇔
L L2 Fu
   
F F F F F
e = 1− − δ⇔e = 1− δ⇔
F0 Fu F0 F0 FR0
F F F
1− +
F0 FR0 F0
δ=e ⇔ (e + δ) = e ⇔
F F
1− 1−
FR0 FR0
F
1−
Fu
(e + δ) = e
F
1−
FR0

The eccentricity at the ends of the column is e. It will be (e + δ) in a deflected

condition at the midpoint, when the column is subjected to the axial load F.

F
1−
Fu
(e + δ) = e
F
1−
FR0

You can then check that the maximum stresses in the cross-sections at the midpoint
and at the top and bottom are not exceeding the ultimate stresses. For the strong
concrete
5.3 Light Concrete Columns and Walls with Entasis 97

F Ec0 F(e + δ)Ec0 hc

+ ≤ fcu
EA0 EI0 2

F Ec0 F(e + δ)Ec0 hc

− ≥ −fctu
EA0 EI0 2

hc is the thickness of the strong concrete. You do the similar check for the light
concrete and for possible reinforcement.
If we know the normal force F and the eccentricity e, we have to design the cross-
section of the column to be able to resist the moment Fe at the end and F(e + δ) at
the midpoint.

5.3.2 Example of a Column with Entasis

In this example, we consider a column with a uniform circular core of strong concrete
of diametre Dc = 160 mm with compressive strength fcc = 45 MPa, tensile strength
fct = 3.8 MPa and initial elastic modulus Ec0 = 36 GPa.
The circular core is surrounded by a circular light aggregate concrete section with
no thickness at top and bottom and thickness 60 mm at the midpoint (Fig. 5.18). This
means that the light combined section has a outer diametre of Dm = 280 mm and an
inner diametre of 160 mm at the midpoint.
The light aggregate concrete has a density of 950 kg/m3 , ultimate strength flcc =
5.0 MPa, tensile strength flct = 0.5 MPa and initial elastic modulus Elc0 = 5.0 GPa.
The column has a height L = 3 m and is loaded by the force F = 300 kN with an
eccentricity e = 20 mm.
π 2 π 4 π π 4
Ac = D , Ic = D , Alc = D2m − Ac , Ilc = D − Ic
4 c 32 c 4 32 m
The ultimate strength of the strong concrete and the total ultimate strength are

Fcc = Ac fcc = 905kN, Fu = Fcc + Alc flcc = 1112 kN

The axial stiffness of the column is

EA0 = Ec0 Ac + Elc0 Alc = 931 MN

The flexural stiffness is

EI0 = Ec0 Ic + Elc0 Ilc = 2506 kN m2

The centrally loaded strength of the column is

98 5 Columns and Walls

8EIc0
F0 = = 2227 kN
L2
The Rankine force of the column becomes
1
FR0 = = 742 kN
1 1
+
Fu F0

The total deflection at the midpoint becomes

F
1−
Fu
(e + δ) = e = 24.5 mm
F
1−
FR0

Checking the stresses at the edges of the light aggregate concrete at the midpoint
you get

FElc0 F(e + δ)Elc0 Dm √

σlc1 = + = 3.67 MPa ≤ flcc = 5.0 MPa
EA0 EI0 2
FElc0 F(e + δ)Elc0 Dm √
σlc2 = − = −0.44 MPa ≥ −flct = −0.5 MPa
EA0 EI0 2

In addition, at the edges of the strong concrete at the midpoint you get
√
σc1 = 20.1 MPa ≤ fcc = 45 MPa
√
σc2 = 3.14 MPa ≥ fct = −3.8 MPa

At the top the strong concrete should carry the load alone, and here we find
√
σc1top = 29.8 MPa ≤ fcc = 45 MPa
√
σc2top = 0.00 MPa ≥ fct = −3.8 MPa

This means that the column can carry the load 300 kN with the eccentricity 20 mm.
We now remove the light aggregate concrete and thereby the entasis and consider
the strong concrete part of the column alone:

π 2 Ec0 Ic 1
FcE = = 1270 kN, FcR = = 528 kN
L 2 1 1
+
Fcc FcE
F
1−
Fcc
(e + δ) = e = 30.9 mm
F
1−
FcR
5.3 Light Concrete Columns and Walls with Entasis 99

And from that

√
σc1m = 38.0 MPa ≤ fcc = 45 MPa
σc2m = −8.15 MPa ≥ fct = −3.8 MPa Failure !

As seen, the strong concrete alone cannot carry the load 300 kN.
It can only carry a load of 233 kN. This is the effect of the Entasis.

References

1. Hertz KD (2019) Design of fire-resistant concrete structures. ICE Publishing, Thomas Telford
Ltd. ISBN: 9780727764447. London, 254p
2. Ritter W (1899) Die Bauweise Hennebique. (In German). Schweizerische Bauzeitung vol 33,
Heft 7, pp 59–61
3. Larsen F (2007) Light concrete structures. MSc Project, Department of Civil Engineering,
Technical University of Denmark. 117p
4. Lind FL (2018) Super-light wall elements. MSc Project, Department of Civil Engineering,
Technical University of Denmark. 82p
5. Vitruvius M (-35) De Architecture. Several publications: Om Arkitektur (In Danish). Translation
edited by J Isager. University of Southern Denmark 2016, 505p
6. CEN (2006) EN 1992-1-2 Eurocode 2, Design of concrete structures, Part 1-2. General rules—
structural fire design. Brussels. 97p
7. CEN (2008) EN 1992-1-1 Eurocode 2, Design of Concrete Structures, Part 1-1. General rules
and rules for buildings. Brussels. 225p
8. The ten books on Architecture (In English). Translated by MH Morgan. Dover 1960. Harvard
University Press 1914, 325p
Chapter 6
Pearl-Chain Structures

Abstract Pearl-chain reinforcement is introduced, where pre-fabricated concrete

segments are post-tensioned together in any desired shape. Pearl-chain arches have
been tested and applied in full-scale, and it is shown how different types of concrete
segments can be used in many variations in roofs, vaults, bridges, cantilevered slabs
etc.

6.1 Pearl-Chains

6.1.1 Pearl-Chain Principle

Arches and vaults are optimal structures capable of carrying loads in pure compres-
sion.
As described in Chap. 1 on history, builders have applied these structures for
thousands of years, and benefitted from large spans, low material consume, elegant
appearance, and no tension that requires expensive materials.
However, it takes labour to construct them. Usually you need a curved scaffolding
and a curved mould for erection of an arch or vault, and labour has become expensive
in the modern society. Consequently, many engineers have not applied these struc-
tures for half a century. In addition, materials applied for scaffolding and moulds
represent extra consume of resources and extra impact on the climate.
Nevertheless, arches and vaults are still geometrically optimal structures with
respect to resisting distributed loads. In many cases they represent the best and most
economical solutions with the smallest material consume and climate impact if they
could only be built without application of scaffolding or curved moulds.
This is the reason why, the first author has invented the pearl-chain principle [1].
By means of this, you can create any curved arch or vault structure from a series
of elements without application of scaffolding.
The single elements can often be equal or consist of very few types and they
can be mass-produced. In many cases, the elements can even be straight or plane
and therefore easy to produce and easy to transport from factory to building site as
demonstrated by the test pearl-chain arch shown in Fig. 6.1.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 101
K. D. Hertz and P. Halding., Sustainable Light Concrete Structures, Springer Tracts
in Civil Engineering, https://doi.org/10.1007/978-3-030-80500-5_6
102 6 Pearl-Chain Structures

Fig. 6.1 Pearl-chain test arch. Photo KD Hertz

The idea came from a little toy animal as the one shown in Fig. 6.2.
When you put pearls on a string and pull the string, you introduce tension forces
in the string and compression forces in the pearls. The compression forces have the
same centroid as the tension forces in the string.
Because you have placed the string in tubular holes in the pearls you can place
the compression forces, where you want them, for example at the centroid of the
cross-section of the pearls.
This means that if a chain of pearls forms a curved shape due to inclined ends of
the pearls, the tension in the string will preserve the curved shape.
The principle is used in the toy animal, where the animal rises and the pearl-chains
become one rigid unit, when the string is tightened.
You may also apply a curved path of the holes between the ends of each pearl.
By doing so, you can avoid that the string should pass edges at the assemblies. In
addition, you can obtain a curved path of the prestressing force, which may be optimal
for resisting distributed lateral forces on each element.

Fig. 6.2 Pearl-chain toy. Photo KD Hertz

6.1 Pearl-Chains 103

6.1.2 Prestressing

Eugène Freyssinet (Fig. 6.3) invented prestressed concrete structures [2]. His idea
was to tension straight or curved reinforcement cables, wires, or rods placed in ducts
in a structure, and anchor them to the concrete introducing similar compression forces
in the concrete sections.
Again, the tension and the compression forces are equal. Their centroids and
directions are the same—also if the cables are curved. This is because the cable
and the concrete mutually interact with lateral stresses changing the direction of the
tension as well as the compression force.
In mass-produced concrete elements such as the SL-deck, prestressing is applied
as pre-tensioned strands. This means that you tension a number of prestressing strands
in a prestressing bed (a mould), before pouring the concrete. Each strand usually
consist of a single wire, and the prestressing bed is often 100 or 150 m long. You
then cut the strands between the elements before they are de-moulded. This process
transfers the prestressing forces to the concrete of each element.
In general, this production method requires some large foundations to resist the
prestressing force during casting, and therefore you mainly produce pre-tensioned
concrete at factories and mainly as straight pre-tensioned strands.
For in-situ cast structures post-tensioned strands are applied. Here, you cast ducts
into the concrete. After casting, you pull strands consisting of a number of wires
through the ducts and tension and anchor them to the ends of the concrete. This gives
good opportunities for designing prestressing strands with curved geometries that
can counteract actual moment and shear distributions [3].
Figure 6.4 shows an example of this. The first author has designed a 40 m long
half-cylindrical shell supported on three walls to carry the load of a roof to both sides.
It is provided with six post-tensioned strands in each side with a shape of a double
suspension bridge.

Fig. 6.3 Eugène Freyssinet.

Drawing KD Hertz
104 6 Pearl-Chain Structures

Fig. 6.4 Post-tensioned half-cylindrical shell. Photo KD Hertz

The structure can resist tension as unloaded compression along these strands.
By means of that, moments and shear from a maximum service load without safety
factors (service limit state) will not give rise to cracking. Additional slack reinforce-
ment at the bottom of the shell resists the difference in force between the ultimate
load (including safety factors) and the service load.
This kind of design, where you apply prestressing to avoid cracking and deflection
for service load and allow cracking for ultimate load is known as partial prestressing
[4–6].
In the half-cylindrical shell, the engineer used prestressing not only to counteract
deflections, but mainly to avoid cracking for shear forces in the slender sides of the
shell.
100 years of experience with prestressing technology means that builders and
structural engineers are familiar with the technology required to pull strands through
a concrete structure and introduce a desired compression.
The benefits of prestressing are well known:
(1) To resist tension as unloaded compression.
(2) To avoid cracking.
(3) To keep the full stiffness of the cross-section when loaded in tension or by a
moment.
(4) To counteract deflection by application of curved or eccentric prestressing.
When using prestressing techniques for establishing a pearl-chain structure of
several concrete elements, you obtain the same benefits.
In addition, you can obtain the following main advantages for curved structures:
(1) To avoid expensive curved moulds.
(2) To avoid expensive curved scaffoldings.
6.1 Pearl-Chains 105

(3) To apply inexpensive prefabricated elements.

(4) To apply equal straight elements, which are inexpensive to produce and
transport.

6.1.3 Pearl-Chain Elements

Figure 6.5 shows how a curved structure can be established using straight elements
with a small inclination of the ends. Each element can contain one or more ducts for
the prestressing lines. The ducts should usually be perpendicular to each inclined
end surface in order to ensure a continuous shape, which is necessary when pulling
the strands.
You would often place the centroid of the ducts centrally in the cross-section of the
element. This gives a uniformly distributed compressive prestress at the cross-section
by preventing bending moments.
If the pearl-chain is an arch, the prestress will be supplemented with the compres-
sive stresses from the main load of the arch. On top of these, you can then add or
subtract bending stresses from secondary skew loads on the arch. Tension is then
resisted as unloaded compression. This means that neither the cross-sections of the
elements nor the joints between them may crack [7].
In case you for some reason decide to make a pearl-chain that does not function
as a perfect arch, which is shaped to support the main loads in compression, you
may compensate for this by placing the prestressing ducts with an eccentricity in the
pearl-chain elements. This is similar to what you do in other post-tensioned structures
as explained in Sect. 6.1.2.
Even for a pearl-chain, which is suppossed to act as a perfect compression arch,
you may want to counteract temporary moments, due to lifting.

Fig. 6.5 Curved pearl-chain made from straight elements

106 6 Pearl-Chain Structures

Fig. 6.6 Tension pearl-chain

The pearl-chain principle is convenient to create curved structures of concrete

easily at a low cost, and with a low environmental impact. It is even more benefi-
cial to apply pearl-chain arches with super-light cross-sections of strong concrete
embedded, stabilized, and protected by light concrete.
By combining super-light structures [8] and the pearl-chain principle [1], you
have the best possible opportunities for placing strong concrete exactly, where you
want the forces to be, and by that design your structure becomes optimal and more
economical.
Prestressing of light concrete has seldom been possible, because the light concrete
may creep when exposed to prestress. By means of super-light pearl-chains, you can
embed prestressed load-bearing parts of strong concrete and at the same time create
a light prestressed concrete structure.
Figure 6.6 is a special example of this. Here a light concrete wall is hung up in a
tension pearl-chain, where tension is resisted as unloaded compression. The figure
shows the pearl chain supported on two columns.

6.2 Pearl-Chain Applications

6.2.1 Pearl-Chain Arches

Figure 6.7 shows two flat super-light deck elements prepared for the first full-scale
test of a pearl-chain arch. The end surfaces are cast with an inclination of 3.5° and a
retarder is applied to the end moulds so that the aggregates are visible. This creates a
rough surface that can transfer shear from one element to another via mortar joints.
In addition, two recesses are made to assist transferring shear between the elements
[9–11].
As a standard, we have designed SL-deck elements with spacing between the
light concrete blocks at the centre. This spacing is designed to allow room for a
post-tensioning cable for pearl-chain structures (see Sect. 4.1.2), and we therefore
positioned the cable duct here. The duct consists of a standard flexible metal tube,
where the prestressing wires are to be pulled. The tube is placed perpendicular to
6.2 Pearl-Chain Applications 107

Fig. 6.7 SL-deck pearl-chain elements. Photo Abeo Ltd.

the inclined end surfaces of the element [12]. This position was simply obtained
by placing a single piece of wood under the middle of the tube before casting the
element. The stiffness of the tube then provided a perfect curve between the ends.
Figure 6.8 shows the first full-scale pearl-chain arch lifted in the middle element
for negative moments and Fig. 6.9 the same subjected to positive moments.

Fig. 6.8 First full-scale test of a pearl-chain arch. Photo KD Hertz

108 6 Pearl-Chain Structures

Fig. 6.9 Load test of first pearl-chain arch. Photo KD Hertz

6.2.2 Integrated Pearl-Chains

Designing two-dimensional constructions like plates, slabs, vaults, and shells, you
may appreciate an opportunity to concentrate forces in arches, ribs, or beams in a
pearl-chain as an integrated part of the structure.
You can apply such a pearl-chain in a larger construction as a skeleton resisting the
main forces. It can also serve to support moulds for the light concrete in a super-light
structure. Often, such integrated pearl-chains are preferable to apply in edges of a
plate, slab, vault, or shell to make the edges robust and to ensure the strong parts the
largest possible lever arm.
If you divide the larger construction into structural elements, you can include the
segments of the pearl-chain as parts of larger elements as shown in Fig. 6.10.
In many cases, integrated pearl-chains serve as arches, ribs, or straight stringer
components transferring axial loads as compression or tension. The pearl-chain

Fig. 6.10 Integrated

pearl-chain
6.2 Pearl-Chain Applications 109

Fig. 6.11 Building with ribs and balloon moulds

prestressing allows you to resist tension as unloaded compression and thereby avoid
cracks and benefit from the full axial stiffness of the cross-section of the pearl-chain.
In addition, you have the same benefits, when you design a pearl-chain for
secondary moment loads, which may appear for example due to skew live loading.
This means that you can apply integrated pearl-chains as ribs of strong concrete
embedded in a light concrete or you can make the rib structure thicker and visible
from the inside with geometry like the one of classical rib construction known from
many cathedrals.
To do so, you may for example establish a system of ribs as pearl-chains of strong
concrete and between them cast a shell of light concrete in situ. Since the light
concrete has an about four times smaller density than the strong (and thereby a four
times smaller mould pressure), the moulds for casting it can be made light, simple,
and inexpensive by means of for instance textiles, which even can be supported by
balloons as shown in Fig. 6.11.
A method for such construction could be:
(1) Build the rib structure.
(2) Blow up balloons between the ribs, where the balloon pressure ensures that the
moulds can fit geometrical irregularities and make the mould tight.
(3) Cast the light concrete between the ribs.
(4) Remove the balloons, and you can benefit from nice vaulted bottom surfaces
of the light concrete between the ribs.
(5) You may cast a thin layer of strong concrete above the ribs, in case the light
concrete is not sufficient to transfer the shear forces.
Alternatively, you can create a ribbed structure of purely factory made elements.
A fully prefabricated design could be to create a frame of strong concrete partly
or fully filled out with a light concrete plate or vault as shown in Fig. 6.12.
Each rib then consists of frame sides from two neighbour elements with a joint
in full height of the rib that is convenient to cast out.
110 6 Pearl-Chain Structures

Fig. 6.12 Two dimensional elements for a ribbed pearl-chain structure

You can include ducts for prestressing in these edge frames, and this allows you
to post-tension a series of such cassette elements as a pearl-chain to be lifted in place
as a whole and to function as a whole in the final structure.
In addition, you may provide the transverse parts of the frames with ducts allowing
you to post-tension the elements across the direction of the primary pearl-chain.
This is relevant for cupolas designed as radial pearl-chains of trapezoidal elements.
Here, you usually require a transverse post-tensioning of the pearl-chains in order
to provide a ring compression force that can counteract the lateral forces from the
cupola.
See more about cupola design in Sect. 8.1.

6.2.3 Cross-Stringer Structures

A cross-stringer structure consists of straight integrated stringers in two directions

connected by shear plates. This structure has proved to be effective as a model for
designing shear walls since the first author developed it in 1976 [13, 14].
For massive walls, you can postulate a distribution of forces in vertical and hori-
zontal stringers. You may choose the stiffness of a stringer to that of the wall next to it
as the stiffness of the total cross-section if the wall is in compression or the stiffness
of the reinforcement if it is cracked. A slack reinforcement stringer can be everything
from a single bar to a bundle of for example nine corrugated bars. The system of
equilibrium equations for shear and stringer forces gives a ribbon matrix, and as soon
as your computer program has solved it, you can easily check the dimensions.
Figure 6.13 shows an example of a folded wall. The stringers crossing the fold
have zero stiffness and the same length on both sides of the fold. This ensures that
only shear is transferred across the fold.
Figure 6.14 shows a cross-stringer model of a wall that acts as a high continuous
beam and the long horizontal openings meant a special challenge for the design.
6.2 Pearl-Chain Applications 111

Fig. 6.13 Stringer model of

a folded wall

Fig. 6.14 Stringer model of a complicated wall

6.2.4 Integrated Edge-Beams

You may often need a one-dimensional structure at the edge of a two-dimensional

wall, slab or shell—capable of concentrating forces or supporting the edge in bending
either in the plane of the structure or out of it.
Figure 6.15 shows such 30 m long edge-beam designed by the first author.
112 6 Pearl-Chain Structures

Fig. 6.15 Prestressed beam at the Royal Danish Theatre in Copenhagen. Photo KD Hertz

It has a cross-section of 1 × 2 m, and is prestressed with eight cables. The curvature

of the prestressing cables is sketched on the photograph. The idea was to counteract
moments from uniformly distributed load of a 100 year old masonry facade above,
before the 18 m wide opening was made below the beam. Tension is here resisted as
unloaded compression in the concrete.
The designer divided the beam into 11 sections in order to cast it stepwise in the
old masonry without challenging the stability of the wall. The beam is an example of a
straight pearl-chain, where the post-tensioning force presses the sections together and
improves the transfer of shear between them in addition to counteracting a moment
distribution.
In the case example, the contractor made an error during construction. Over two
sections of the beam next to the middle section, he placed the prestressing ducts in a
straight line, because he misunderstood the drawing and read numbers for length as
numbers for height.
The young engineer had to consider this overnight, because it would be costly
and time consuming, if the contractor should remove the two sections and cast them
again.
He concluded that this rather serious error did not mean much to the function of
the beam, since the resulting lateral force depends on the difference in inclination
between the ends and not on the shape of the curve.
Therefore, as long as the curve is smooth and the beam can resist the secondary
moments from a less optimal path, the detailed shape is less important.
Another mishap at this particular structure gave rise to good learning as well.
Something hindered pulling one of the cables through its duct. Some cement paste
had penetrated one of the duct joints between two sections.
6.2 Pearl-Chain Applications 113

Fig. 6.16 View inside a duct

with penetrating cement.
Photo KD Hertz

The solution was to call a sewer contractor, who could guide a small rolling robot
with a lamp, a camera, and a small drill with a fan into the duct removing the cement
paste. Figure 6.16 shows a picture taken by the robot. The whole operation only
caused a six hours delay.
In some cases, you may be interested in application of an embedded integrated
pearl-chain beam designed to carry the load of adjacent slab elements between
supporting columns.
If the integrated transverse part of the pearl-chain beam has the same thickness
as the slab, you can avoid the hindrances, which supporting beams usually make
for design. For instance with respect to reduction of natural light, increasing storey
height, and obstructing building services.
As an example, we can consider an edge-beam integrated across the ends of a
number of SL-deck elements. The intention of the beam is to prestress the SL-deck
ends across the direction of the span to form a straight pearl-chain beam that can
support the deck between columns in a facade.
Consider for example a long office building with SL-decks spanning across the
building from facade to facade. We can provide the SL-decks with cable ducts in
both ends and create a post-tensioned pearl-chain edge-beam in each facade.
The edge-beam has the purpose of making the position of the columns supporting
the deck independent of the element division.
The erection procedure could be as follows: At first, a temporary steel C-profile
is mounted on the columns to support the ends of the deck elements.
Then, we pull prestressing cables through the ducts in the elements in the length
of the facade and post-tension the deck element ends together as a pearl-chain. Then,
we can remove the temporary steel profile, and the edge beams carry the load of the
deck structure to the columns.
In the example, we used a span between the columns equal to the width of three
deck elements. If we could make the slab elements fixed end supported, their span
could be increased 50%. We therefore designed the beam to create a fixed end support.
Likewise, we prestressed the pearl-chain beam to be continuous and fixed end
supported over the columns.
114 6 Pearl-Chain Structures

Fig. 6.17 Integrated edge beam test element design [15]

This means that we had to design the pearl-chain edge-beam to resist bending and
shear between the columns as fixed end beams and in addition to resist the torsion
from the fixed end support of the ends of the SL-deck elements.
A structure like this has been tested for ultimate load in bending, shear, and torsion
at the structural lab of the Technical University of Denmark by Iskau and Nissen [15].
Three specimens were tested each of them consisted of three SL-deck elements,
which were shortened to be able to fit into the test space. Each deck element was
2.4 m wide and the edge-beam was post-tensioned with two cables.
Figure 6.17 shows the design of one deck element end including reinforcement
of the integrated pearl-chain edge beam.
Figure 6.18 shows the test setup, where two hydraulic jacks to the right apply
forces to the central of the three deck elements creating a simultaneous bending,
shear, and torsion in the edge-beam.
The main conclusion was that the full-scale tests did not show any unforeseen
matters, and that you will be able to design a post-tensioned pearl-chain edge beam
and utilize its load-bearing capacity by means of common static principles.

6.2.5 Tension Ties

Pearl chains may serve as tension ties, where you can avoid cracks in a tension zone
and benefit from the large stiffness of the unloaded concrete section compared to that
of the prestressing steel itself.
Tension ties can be separate pearl-chains or imbedded in a number of elements.
6.2 Pearl-Chain Applications 115

Fig. 6.18 Edge beam test setup. Photo KD Hertz

For example, you can apply a vertical post-tensioning of both edges of a cross-
wall of a high-rise building in order to resist bending from wind load with maximum
stiffness and without formation of horizontal cracks as shown in Fig. 6.19a.
Figure 6.19b shows another example.
A continuous beam supports a deck over the ground floor of a theatre. A row
of columns supports the beam. However, one of the columns would hinder turning

Fig. 6.19 Vertical pearl-chain tension ties a shear walls b hung up beam
116 6 Pearl-Chain Structures

Fig. 6.20 Vault supported by a post-tensioned floor slabs b drilled tension ties

of thirty metre long carpet roles into a carpet elevator. We solved the problem by
designing an integrated tension tie through the edge of the elements of a shear wall
above, so that the shear wall could carry the beam in this particular point.
Engineers making arches, vaults, and arch bridges often consider lateral forces to
represent a problem that gives rise to design of large foundation blocks. Especially,
when you design a flat arch, the lateral forces may become substantial.
In many cases, you can apply the classical solution to this problem by establishing
a tension tie between the supports reducing the reaction of the entire structure to be
vertical forces.
Designing a building, you may guide the inclined forces of a vault down to the
floor slab below. You could do this by supporting the edges of the vault with edge-
beams supported by a number of shear walls taking the reactions to the floor as shown
in Fig. 6.20a.
You could then post tension the floor slab to take the horizontal components as
unloaded compression in the deck elements as a tension tie pearl-chain. In this way,
you could create a vault at any floor of a building, even the top floor of high-rise
buildings, only resulting in vertical forces on the structures below.

6.2.6 Drilled Tension Ties

In case of a highway bridge, you could sometimes create a similar post-tensioned

floor slab as part of the road below. However, if the highway is in service, it will
cause a stop of traffic, if you should do this. For a situation like this, you could design
a number of drilled tension ties (Fig. 6.20b).
Many contractors are familiar with a technology of drilling a horizontal tube for
example with a diameter of 600 mm under a highway while traffic is running.
You could then design a number of these drilled tubes filled with concrete either
in-situ cast or as pre-cast segments. You provide each of these horizontal columns
with one or more cable ducts and post-tension them.
You have now made a series of horizontal tension ties of prestressed concrete that
can resist lateral forces from the arch bridge superstructure as unloaded compression
with the benefit of the large stiffness of the concrete cross-section. By means of this,
6.2 Pearl-Chain Applications 117

it is possible to build a highway bridge with a minimum of traffic disturbance. Only

for a few hours, the traffic must be stopped while the pearl-chain arches are lifted in
place.

References

1. Hertz KD (2008) Light-weight load-bearing structures reinforced by core elements made of

segments and a method of casting such structures. (Patent on Pearl-Chain technology). Europe
EP 08160304.5, USA US 61/080445, PCT (Patent Cooperation Treaty) PCT/EP2009/052987,
13. Patent 2012
2. Freyssinet E (1936) Progrès pratiques des méthodes de traitement méchanique des beétones.
(Praktische Weiterentwicklung der Verfahren zur mechanischen Behandlung von Beton, Prac-
tical Improvements in the Mechanical Treatment of Concrete), Zweiter Kongress der IVBH =
Deuxième Congrès de l’AIPC = Second Congress of IABSE, Berlin-München 1.-11. October
1936. IIb3 pp 205–231
3. Leonhardt F (1964) Prestressed concrete. Design and construction. Ernst W, Sohn, 2nd edn,
677p
4. Brøndum-Nielsen T (1972) Concrete structures II. (In Danish). Structural Research Laboratory,
Technical University of Denmark, 125p
5. Brøndum-Nielsen T (1973) Structural concrete. Technical University of Denmark, Structural
Research Laboratory, p 136
6. Brøndum-Nielsen T (1976) Partial prestressing. Report R 76, Structural Research Laboratory,
Technical University of Denmark, 26p
7. Halding PS (2016) Construction and design of post-tensioned pearl-chain bridges using SL-
technology. PhD Thesis. Report R-350, Department of Civil Engineering, Technical University
of Denmark, 204p
8. Hertz KD (2007) Light-weight load-bearing structures (patent on super-light struc-
tures). Europe EP 07388085.8, USA US 61/004278, PCT (Patent Cooperation Treaty)
PCT/EP2008/066013, 21. Patent 2010
9. Hertz KD, Halding PS (2014) Super-light pearl-chain arch vaults. In: Paper nr. 52 in proceedings
of the IASS-SLTE 2014 symposium, “shells, membranes and spatial structures”, 8p September
2014, Brasilia, Brazil
10. Lund MSM, Hansen KK, Hertz KD (2016) Experimental investigation of different materials in
arch bridges with particular focus on pearl-chain bridges. Construct Build Mater 124:922–936.
(Elsevier August 2016)
11. Lund MS (2016) Durability of meterials in pearl-chain bridges. Ph.D. thesis. Report R-341,
Department of Civil Engineering, Technical University of Denmark, 157p
12. Halding PS, Hertz KD, Viebæk NE, Kennedy B (2015) Assembly and lifting of pearl-chain
arches. Paper 71 of fib symposium concrete-innovation and design, 10p. Proceedings p 185.
Copenhagen May 2015
13. Hertz KD (1976) The Stringer method applied for calculational design of building structures.
(In Danish). MSc project. Department of Building Design, Technical University of Denmark,
147p
14. Hertz KD (1978) Calculation of shear plate structures by means of crossing stringer systems.
(In Danish). Bygningsstatiske Meddelelser vol 49, no.4. pp 113–129.
15. Iskau MR, Nissen JS (2016) Design and full-scale test of integrated beam in SL-decks. MSc
thesis Department of Civil Engineering, Technical University of Denmark, 256p
Chapter 7
Arch Bridges and Vaults

Abstract Arches and vaults are minimal structures, and basic calculation methods
are provided for them. Pearl-chain bridges and vaults are introduced, and a bridge is
build based on the technology. The chapter also introduces sandwich arch bridges,
which is a method to upgrade traditional arch bridges with increase of the capacity.

7.1 Arches

7.1.1 Arch Principle

Arches are optimal structures capable of carrying uniformly distributed gravity loads
in pure compression (without bending moments).
Figure 7.1, shows how pearl-chain structures provide new possibilities for making
arches simple, inexpensive, and fast from flat mass-produced deck elements.
As seen from Figs. 1.7, 1.9, and 1.10, builders have applied arches for thousands
of years for vaults in buildings and for arch bridges. Some of these arch structures
were in service for millennia.
In general, the advantage of arches is that you can carry a uniform load by a
minimal structure made of inexpensive and sustainable materials, which are mainly
suitable for taking compression such as stone, mortar, and concrete. Furthermore,
you can avoid the expensive and less sustainable materials to resist tension.
Figure 7.2 shows a comparison between a possible force distribution in an arch
and in a beam with fixed ends, uniformly distributed load p and span L. Both get
the same moments if they are simply supported. A beam with fixed end supports
get reduced moments, but the difference between maximum and minimum moments
is unchanged. The arch can totally avoid moments for a uniformly distributed load
by fixing the supports for lateral movements. You find the lateral arch force from a
condition of no moment at the midpoint.
If you have live load on a bridge or a vault, you cannot avoid bending moments.
A single load P is often critical in a quarter point of the span at L/4 from the one
end at a height h. It gives vertical reactions of P/4 and 3P/4. If you consider your
arch to have a hinge in the middle, you can take moment about that, and you get a

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 119
K. D. Hertz and P. Halding., Sustainable Light Concrete Structures, Springer Tracts
in Civil Engineering, https://doi.org/10.1007/978-3-030-80500-5_7
120 7 Arch Bridges and Vaults

Fig. 7.1 Pearl-chain arch bridge across Vorgod River, Denmark. Photo Abeo Ltd.

Fig. 7.2 Arch and beam statics

lateral reaction force of size PL/8H and the moments in the quarter points at height
h are then PL(3/16 – h/8H) and PL(1/16 – h/8H).
However, dead load is often a relatively large part of the total loading. The large
normal force in the arch gives you compressive stresses, which in traditional stone
arches allows adoption of tensile stresses from secondary moments as an unloading
of the compression.
7.1 Arches 121

7.1.2 Arch Shapes

Which curve is then the optimal for the dead load? This depends on the load on the
arch. If you have a vertical load, which is uniformly distributed over the span, the
optimum is a parabola. If the arch must only carry its own weight, the optimum is a
chain line also called a catenary. If you have both, the optimum can be a circle for
flat arches, because it will be close to the average of the other two shapes.
If the arch span is for example eight times its height, the ratio between the height
in the quarter point and the arch height at mid-span will be 0.750 for the parabola,
0.765 for the catenary, and 0.761 for the circle and the curves looks as shown in
Fig. 7.3 at the top.
When the arch span is four times its height, the ratio between the height in the
quarter point and the arch height at mid-span will still be 0.750 for the parabola,
0.765 for the catenary, and now 0.791 for the circle and the curves looks as shown
in Fig. 7.3 at the bottom.
If you make arches and vaults with circular curves, you have the big advantage
of being able to build them from equal elements. As you see, the circular arch is in
general the most optimal curve to apply for relatively flat structures.
Builders have therefore applied circular arches to support walls over openings for
windows, doors, and gates and to make daring bridges (Fig. 1.10).
By increasing the width of the arches, you get circular vaults suitable as roof
structures (Fig. 1.9) or, ceilings in large rooms supporting the upper stories.
In tunnels and sewers, builders have used circular vaults to sustain the earth
pressure and in many cases, they have made the cross-section a full circle since their
structure should not only receive forces from above, but also deliver them to the
underlying material and sustain pressure from the sides.
Designers have not only applied circular arches and vaults to sustain vertical
forces, they have also e.g. made horizontal compression rings around holes in the top

Fig. 7.3 Arch curves: Parabola ypa , catenary yca , and circle yci for H/L = 1/8 and 1/4
122 7 Arch Bridges and Vaults

of cupolas (Fig. 1.14), and horizontal arches to support distributed lateral pressure
from torus shells as in Fig. 1.11, from soil as in Fig. 1.12, or from water in dams.

7.2 Pearl-Chain Arches

7.2.1 Pearl-Chain Arch Bridges

Builders have always had the main disadvantage that construction of a compression
arch required a temporary scaffolding resisting the loads until the arch was completed.
The pearl-chain principle allows you to build arches without scaffolding and
thereby make arches more applicable. You can especially make a circular pearl-
chain arch simple, inexpensive, and fast, because you can design it from equal, flat,
and mass-produced elements.
In order to avoid all expensive curved supports while assembling pearl-chain
arches, the authors have found that a successful method is to place the elements in a
vertical position on a plane surface before post-tensioning them together [1, 2]
In Fig. 7.4, a pearl-chain arch is lifted by crane after the assembly.
Prior to the post-tensioning, the joints of width 2.4 m were cast out in the vertical
position. The method for pouring the joints were tested in 2015. It was proven that
a mortar could be used to fill out the joints sufficiently [3, 4].
The vertical assembling method allows a simple procedure of erection.
Flat SL-deck elements are transported to a place near the building site, which for
a freeway bridge may be at the roadside.

Fig. 7.4 Pearl-chain arch after assembling. Photo KD Hertz

7.2 Pearl-Chain Arches 123

Fig. 7.5 Full-scale test of pearl-chain bridge for skew load. Photo KD Hertz

Here, the contractor assembles all arches and casts out the joints before a
specialized crew arrive to prestress the pearl-chain structures with post tension cables.
After that, you can lift all arches in place quickly without disturbing the traffic
more than about 4–5 h for a typical freeway bridge.
Before the first pearl-chain arch bridge was established, we made a full-scale
test for skew load. It was a concentrated load in one of the quarter points, which as
explained in 7.1.1 traditionally is the most dangerous load case for an arch bridge.
Figure 7.5 shows the test setup. It proved to sustain almost the double load, for
which we had conservatively designed it [5–7].
Figure 7.6 shows a similar test made in 1893 before a Monier arch was applied
by Asger Ostenfeld in the Gefion Bridge in Copenhagen in 1894 (Fig. 7.7).
The bridge stood 104 years until the community would like to change its position.

7.2.2 Long Span Pearl-Chain Arch Bridges

Sometimes engineers would like to build an arch bridge with a long span.
Of course, they would like to do that without application of a big costly curved
scaffolding.
Gustave Eiffel gave a solution to this problem. In 1877, he constructed the Maria
Pia Railway Bridge across the Douro River in Porto with a single steel lattice arch
spanning 160 m and with a height of 60 m.
Figure 7.8 shows his solution. He prolonged the two towers next to the basements
of the main arch with temporary towers between which he made a cable like that of
a suspension bridge. He applied this cable to lift parts in place and combined with
124 7 Arch Bridges and Vaults

Fig. 7.6 Monier arch test for skew load 1893 before building the Gefion Bridge

Fig. 7.7 Gefion Bridge with Monier arch from 1894. Photo KD Hertz

inclined cable stays, he could support parts of the arch before it was completed and
could carry itself.
The same procedure could be applied for construction of a pearl-chain arch with a
long span. The arch could be constructed by two half pearl-chain arches or of smaller
segments kept in place and temporary supported by a main cable between temporary
towers and by inclined cables to the sides.
7.2 Pearl-Chain Arches 125

Fig. 7.8 Maria Pia Bridge. Drawing KD Hertz

For small flat arch bridges, you can transfer the distributed load from the road or
rail to the arch by means of a filling material. For large arch bridges, you may have
to apply intermediate columns. The columns act with concentrated loads on single
points of the arch.
This means that you do not need the curvature of the arch in between the points,
where it supports the columns for anything else but the dead load of the arch itself.
You can therefore often make a more optimal pearl-chain arch using straight
elements connected at angles at the points, where it supports the columns. You may
perhaps include special elements at the support points providing the angles and
supporting the columns.

7.2.3 Half-Arch Structures

In many cases, you can take lateral forces from arches and vaults in an even more
simple way by application of half-arches. We originally developed the method
for bridges crossing streams, where tension ties between the supports would be
impossible, and where soft soil meant that pile foundations could not resist lateral
forces.
At each side, you supplement the arch with a half-arch that will provide the same
lateral pressure as the main arch. See Fig. 7.9. You counteract the normal force of
the two half arches by a tension force in a tie above the arches. The tie may be a
post-tensioned deck capable of resisting tension as unloaded compression.
You can consider the structure as consisting of four half arches or of two sections
with cantilevered half arches to each side counterbalancing each other.
A series of these types of super-light structures were already tested to failure by
Larsen [8] before the pearl-chain principle was invented (Fig. 7.10).
Two half arches were cast in a curved mould and embedded in light aggregate
concrete. A tension tie in the top transferred its force to the compression arch with
a steel plate at each end. At failure load, the strong arch crushed, where its end
penetrated from the light concrete and was loaded by the steel plate (Fig. 7.11).
126 7 Arch Bridges and Vaults

Fig. 7.9 Half-arch solution with horizontal tension tie

Fig. 7.10 Test of super-light double half arch. Photo KD Hertz

The half-arch structure is suitable for creating large vaulted spaces at any storey
in a building, because it only transfers vertical loads to the lower levels. The outer
half arches allow architects to design open facades without columns. This is possible,
since a cantilevered half arch can carry vertical loads.
Half-arch structures are also suitable for designing bridges over streams, roads,
and rails, where the outer half arches fit to the sloping embankments that you often
find to the sides of the highway lanes or rail tracks. In Sect. 7.3.4 it is shown, how
you may shorten the outer half arches if required.
7.3 Pearl-Chain Sandwich Arches and Vaults 127

Fig. 7.11 Fracture of the end of a strong arch core. Photo KD Hertz

7.3 Pearl-Chain Sandwich Arches and Vaults

7.3.1 Pearl-Chain Sandwich Arches

Traditional arch builders have faced two main problems.

(1) A requirement for costly curved scaffolding.
(2) A requirement for heavy foundations to resist the lateral forces.
(3) A requirement for a an arch cross-section that can resist secondary moments.
The first problem is solved by application of pearl-chains introduced in Chapter 6
and in Sect. 7.2.
The second problem can be solved by application of post-tensioned ties as shown
in 6.2.5 and in 6.2.6 and by application of half arches as shown in 7.2.3.
The third problem can be solved by application of what we call sandwich arches.
Skew loading causes the secondary moments, typically from a concentrated load
close to the quarter point.
Figure 7.12 shows Ponte de Mosteiró constructed by Prof. Edgar Cardoso in 1972
across the Douro River. It has a span of 110 m and its arch height is only 7.2 m. It is
an example of a half arch structure, as introduced in 7.2.3.
It is also an example of a bridge, where a lattice—here an X lattice—is capable
of transferring shear between arch and top slab, so that a pair of forces can resist
secondary moments from live load with the entire lattice height as lever arm.
Eugène Freyssinet applied this trick first in the Boutiron Bridge from 1912 across
the river Allier near Vichy. In this bridge he applied a V-lattice to transfer the shear
and he established a span 72 m with a height of 5.2 m.
128 7 Arch Bridges and Vaults

Fig. 7.12 Ponte de Mosteiró. Photo KD Hertz

This appears to represent a general solution to the old problem of secondary

moments from skew loads on arch bridges. For smaller arches, it will not pay back to
build a lattice. Here, the obvious solution is to establish a layer of material that can
transfer shear between the arch and a top plate as shown in Fig. 7.13. The top plate
can serve as basis for a rail or a road pavement, or can function as the pavement.
The cast shear transferring material could for example be a light pervious concrete
with a modest tensile strength of a few MPa. Such a pervious concrete could ensure
that penetrating rain is guided to the top of the arch and away from the bridge.
For positive moments with compression at the top, you may then resist the tension
by unloading the compression force in the arch.
In case you would avoid cracks in the top plate for negative moments, you could
decide to post-tension the top plate after it has been cast in situ or placed as a layer
of concrete elements.

Fig. 7.13 Principle of a sandwich arch

7.3 Pearl-Chain Sandwich Arches and Vaults 129

Fig. 7.14 Test of a scaled sandwich arch bridge

Figure 7.14 shows one of a series of model tests made with the sandwich arch
principle in order to unveil, if anything should hinder a solution like that.
Nanna Tange Bech and Maria Holst [9] created a series of scale model SL-deck
elements in strong concrete and with small blocks of model light aggregate concrete.
They then prestressed these model SL-deck elements to a number of pearl-chain
arches.
A light filling was then poured on top of the arch, and they finished by making a
series of scaled massive deck elements and prestressed them to become pearl-chain
top plates.
They tested the arches with varying stiffness of the filling between arch and top
plate.
The tests were performed with a point load in the quarter point.
The scaled tests demonstrated that the sandwich arch principle can increase the
load-bearing capacity with respect to secondary moments from skew load more than
ten times.
Figures 7.15 and 7.16 illustrates a simple version of the sandwich arch, where
a pervious light concrete is cast out between a pearl-chain arch and an in-situ cast
top slab. Here, the sandwich layer has simple inclined sides. This was applied in
the project shown in Fig. 7.15 for a pedestrian bridge designed by Henning Larsen
Architects Ltd from 2015.

7.3.2 Sandwich Arch Bridges

Figure 7.17 shows a bridge designed and built in 2015 across Vorgod River by Sweco
Engineers Ltd and Abeo Ltd. It has a main arch with a span of 13 m and two half
arches of 6.5 m each.
130 7 Arch Bridges and Vaults

Fig. 7.15 Pearl-chain Sandwich arch bridge project. Henning Larsen Architects Ltd.

Fig. 7.16 Simple sandwich arch principle

Fig. 7.17 Super-light pearl-chain bridge across Vorgod River

7.3 Pearl-Chain Sandwich Arches and Vaults 131

As described in Sect. 7.2.3 the half-arch structure is capable of lifting the lateral
forces from the main arch up to be withstood as tension or unloaded compression in
the bridge deck.
Here it is required, and it is a convenient way of building, because the soil is soft
and the bridge is supported on a pile foundation that cannot give sideward resistance.
Figure 7.1 shows the same bridge during construction. At first, the contractor
established the pile foundations. Then, he made a preliminary sideward stabilisation
of them by means of horizontal prestressed Macalloy rods as tension ties.
Next, he assembled the pearl-chain arches and lifted them in place, and the lifting
operation took only five hours.
This demonstrates that you can reduce interruption of traffic to this time, when
you establish a pearl-chain bridge across a freeway.
With the arches in place, a pervious concrete was cast on top of them. It was
developed by Lund [3] in collaboration with researchers at Iowa State University to
be cast in sufficient thicknesses for the purpose. At last, the top slab was cast, and
the preliminary tension ties could be removed.

7.3.3 Cassette Sandwich Arches

We designed the arch in the Vorgod River bridge in Fig. 7.17 as a pearl-chain of
flat SL-deck elements. The SL-deck factory could mass-produce them quickly and
inexpensively as a part of their ongoing deck element production.
However, for an arch in a bridge you do not need all advantages of the SL-decks
such as sound insulation. In many cases, you can therefore save material, money, and
CO2 by application of what we call cassette elements similar to the one introduced
in Fig. 6.12 in Sect. 6.2.2.
A cassette consists of a frame of strong concrete filled with a plate of light concrete
in less height. The outer sides of the frame serve as sides of the joints to the neighbour
elements. In the longitudinal direction, the sides are inclined in order to give the
curvature and shape of the arch. The longitudinal sides of the frame contain tubes
for post-tension cables.
Using such cassette elements, you can benefit from the voids for more than
reducing weight. If you turn them upwards as shown in Fig. 7.18, the shear transfer-
ring filling material can get an effective grip in the top surface of the arch. That will
ensure that there is no slipping between the surface of the arch and the filling.
It is also possible to apply the sandwich principle to ensure lateral stability of
structures consisting of a number of high and slender arches.
In Fig. 7.19 is shown such an example of two adjacent long and slender arches.
It is Ponte da Arrábida, which Edgar Cardoso constructed in 1963 in Porto,
Portugal.
It has a span of 270 m, a height of 70 m and a width of 27 m.
132 7 Arch Bridges and Vaults

Fig. 7.18 Reversed ribbed arch of cassette elements

Fig. 7.19 Ponte da Arrábida. Photo KD Hertz

In order to give them stability for transverse forces, he constructed the arches with
a certain distance and connected them with a horizontal shear transferring X-lattice.
The shear transfer hinders that the arches can move relative to each other in their
longitudinal direction, as they would, if they should tilt sidewards.
Likewise, you may design material saving and slender pearl-chain arches
separated by light concrete sections serving as shear plates.
A special design of such pearl-chain arches for bridges and vaults is by application
of the ribbed cassette elements as shown in Fig. 7.18. Here the sides of the cassettes
constitute the arches, which are separated by the light concrete in the shear plates.
This can create very stable arch structures for bridges and vaults with a minimum
of materials applied for the elements.
7.3 Pearl-Chain Sandwich Arches and Vaults 133

7.3.4 Lattice Effects in Sandwich Arches

If you want to improve the shear capacity of the filling material of a sandwich arch or
perhaps even apply a material that can only resist compression, you could introduce
some vertical or inclined tensile members. For example, such members could be
made as pre-tensioned columns of strong concrete, which can withstand tension as
unloaded compression. They can then transfer shear as tension in the columns and
compression in the filling material as shown at the top of Fig. 7.20.
The figure also shows examples of how you may save filling material in deep
sandwich arches by placing horizontal round or triangular tubes creating cavities
or sand filled volumes around which the filling constitutes a lattice. As shown in
Fig. 1.10, builders have applied such holes since antiquity to save materials and to
unload sidewards pressure from streaming water, when the water level of the river is
high.
Finally, the figure shows how you may reduce the length of outer half arches and
still counterbalance the horizontal arch force by introducing a vertical tension force
for example from a soil anchor like those applied for sheet piling or by a concrete
block or a box with sand.

Fig. 7.20 Lattice formation in sandwich arches

134 7 Arch Bridges and Vaults

Fig. 7.21 Gallery bridge project. Project and rendering by Timo Nielsen et al. 10

7.3.5 Shear Wall Sandwich Arches

Figure 7.21 shows a project for a gallery bridge intended to cross the River Thames
in London by Nielsen et al. [10]. It has a main arch with a span of 176 m and two
half arches each spanning 88 m.
In this project, the sandwich effect is established by means of shear transferring
walls.
The walls make it possible to utilize the space between the arch and the deck of
the bridge as a two storey high gallery building.
For the construction is was suggested to establish a preliminary support at the
middle of the main span. An alternative solution could be two preliminary pylons
over the main foundations in the river with tension cables, as suggested by Gustave
Eiffel and shown in Fig. 7.8.
At the centre span, the project group designed a top arch that extended the space
for the gallery and provided the bridge with a supplementary load-bearing capacity.
Such a double effect is also seen in the Simone de Beauvoir Bridge in Paris from
2004 by Dietmar Feichtinger, and in the Royal Albert Bridge at Plymouth from 1854
by Isambard Kingdom Brunel.

References

1. Halding PS, Hertz KD, Viebæk NE, Kennedy B (2015) Assembly and lifting of pearl-chain
arches. In: Paper 71 of fib symposium concrete-innovation and design, 10p. Proceedings p 185.
Copenhagen May 2015
References 135

2. Halding PS, Hertz KD, Schmidt JW (2015) Precast pearl-chain concrete arch bridges. Eng
Struct 103:214–227. (Elsevier 2015)
3. Lund MS (2016) Durability of meterials in pearl-chain bridges. Ph.D. thesis. Report R-341,
Department of Civil Engineering, Technical University of Denmark. 157p
4. Lund MSM, Hansen KK, Hertz KD (2016) Experimental investigation of different materials in
arch bridges with particular focus on Pearl-Chain Bridges. Construct Build Mater 124:922–936.
(Elsevier August 2016)
5. Halding PS (2016) Construction and design of post-tensioned pearl-chain bridges using
SL-technology. Ph.D. Thesis. Report R-350, Department of Civil Engineering, Technical
University of Denmark, 204p
6. Halding PS, Hertz KD, Schmidt JW, Kennedy B (2017) Full-scale load tests of pearl-chain
arches. Eng Struct 131:101–114
7. Rossi M (2015) Pearl-chain arch bridges above electric railways. MSc thesis, Universitá degli
studi di Genova, 132p
8. Larsen F (2007) Light concrete structures. MSc project, Department of Civil Engineering,
Technical University of Denmark, 117 p
9. Bech NT, Pedersen MH (2015) Development and test of arch bridge with materials of different
stiffness. MSc thesis, Department of Civil Engineering, Technical University of Denmark, 235p
10. Nielsen T, Henriksen T, Bendtsen R, Østerby I, Petersen F (2015) Gallery bridge. Project in
super-light structures. Technical University of Denmark, 31p
Chapter 8
Shells

Abstract Element cupolas are possible to create by using light trapezoidal concrete
elements and pearl-chain reinforcement. A calculation method is provided for
cupolas, and other possible modular shell shapes are introduced together with tubes,
tunnels, and floating Super-light structures.

8.1 Shell Structures

8.1.1 Shell Structures

Shell structures can give powerful architectural expressions and make it possible for
builders to create large spaces in buildings. You can make shells with application
of a minimum of building materials and shells therefore have a huge potential for
sustainable building design (Fig. 8.1).
Unfortunately, architects and engineers do almost not apply shell structures in
modern buildings. This is mainly because shells usually require costly manmade,
customized moulds, complicated casting, and demanding engineering calculations.
However, the technologies dealt with in this book can remove all these problems.
We therefore hope that the great potential of shells will be applied more often to
create functional and expressive structures (Fig. 8.2).

8.1.2 New Possibilities

Pearl-chain technology makes it possible to create inexpensive curved shapes by

straight, mass-produced elements, which can easily be transported and lifted in place.
See Chap. 6.
Super-light mass-produced cassette elements as shown in Fig. 6.12, opens possi-
bilities for creating light curved shell structures, which are inexpensive and gives
a small material consume, -building cost, and -CO2 impact. You can apply textile

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 137
K. D. Hertz and P. Halding., Sustainable Light Concrete Structures, Springer Tracts
in Civil Engineering, https://doi.org/10.1007/978-3-030-80500-5_8
138 8 Shells

Fig. 8.1 Polylithic. A super-light modular shell structure by Peters et al. [1]

Fig. 8.2 Super-light tube bridge by Herdal et al. [2]

moulds due to the small mould pressure of light concrete, and they open new poten-
tials for creating shells with double curved shapes (Sect. 8.6.2). Inflatable moulds
enables builders to create double curved structures and cassette elements with vaulted
fillings of light concrete inside frames of ordinary load-carrying concrete, as for
example shown in Figs. 6.11 and 7.18.
In addition to vaults, which we mainly deal with in Chap. 7, the present chapter
presents:
8.1 Shell Structures 139

• Cupola design with only a few different elements and built without scaffolding,
• Torus structures for round buildings with open facades.
• Modular elements for creating free shapes of shells,
• Hyperbolic shells,
• Tubular structures and tunnels,
• Floating submerged tunnels,
• Structures cast underwater,
• Floating bridge foundations,
• Ships of concrete.

8.2 Cupolas

8.2.1 Cupolas

Builders have used concrete cupolas or dome structures for thousands of years
(Figs. 1.14 and 3.3) to create large indoor spaces. A cupola is a statically inde-
terminate structure, which according to Sect. 3.2 means that you as a builder can
apply direct engineering and determine where the forces should be.
For dead load, which usually is the largest load contribution, and for uniformly
distributed live load, it would be most natural to apply a symmetrical set of compres-
sion forces along the medians. For a flat cupola as the one shown to the left in Fig. 8.3,
equilibrium of the median compression forces and the weight of the elements will
give rise to compressive horizontal ring forces, except at the bottom, where you need
to resist outwards horizontal components of the median compression forces.
You can resist the reaction forces either by a solid foundation or by a ring shaped
tension tie that you can create as a prestressed pearl-chain concrete ring withstanding
the tension as an unloading of the pre-compression.
Piere Luigi Nervi used a post-tensioned ring foundation in his Palazzetto dello
Sport as seen in Fig. 3.3.
At a certain inclination of the cupola, it is not considered flat any more, and ring
tension forces occur in the shell as well. At this specific inclination, Nervi introduced

Fig. 8.3 Cupola and hall

140 8 Shells

glass windows in the Palazzetto and guided the forces along the radians into inclined
columns supported on the prestressed ring foundation. The same is done in the cupola
of Hagia Sophia in Istanbul.

8.2.2 Element Cupolas

You can divide a cupola in halves and insert a cylindrical vault (Chap. 7) as shown
to the right in Fig. 8.3. You may then transfer the horizontal compression- or tension
forces of the two half cupolas by horizontal stringers in the vault. You may transfer
the horizontal lateral forces of the vault itself as unloaded compression in a post-
tensioned floor slab or in a series of post-tensioned concrete rods cast in tubes under
the floor as explained in more detail in Sects. 6.2.5 and 6.2.6.
You can insert windows at the top of the vault by means of the earlier described
cassette elements, where you can exchange the light concrete filling with glass.
You can apply the strong cassette edge profiles as “columns” to transfer compression
forces of the vault in the direction of the arch and as “beams” to transfer compression
forces from the one half cupola to the other in the longitudinal direction.
You can make a cupola of plane mass-produced prefabricated elements. To under-
stand how, you may first consider the well-known construction method of prefabri-
cated slurry tanks. Factories produce such tanks ready-made with diameters up to
about 60 m.
First, you make a ring-shaped foundation. On that, you place a number of equal
plane rectangular concrete panels vertically, with a small angle between them and
then grout the joints. Then you pull circumferential post-tensioning strands around
the tank and tension them with a device such as a DYWIDAG ME Floating Coupler,
where the strands overlap and you stress them by means of a simple belt-buckle
principle.
This means that no anchoring blocks and complicated anchoring reinforcement
is required.
The ring-strands prestress the circle of concrete panels so that they can resist
tension from the pressure of the content of the tank as unloaded compression with
the stiffness of the concrete and without cracking.
Now imagine that you instead of vertical rectangular panels place a circle of
trapezoidal panels with an inclination inwards to the centre of the circle. The geometry
of the elements will determine how much they tilt, since they should lean against
each other as seen in Fig. 8.4. The contact between the inclined elements will give
rise to a lateral compression at the top and tension at the bottom ring foundation.
The ring foundation should therefore be post-tensioned to resist this tension as
unloaded compression. The lateral compression means that the tilting elements are
stable as long as they have adjacent elements on both sides. To give the first tilting
panels an extra support during construction, you may place a pole at the centre from
which wires can fix upper corners of the panels as also shown in Fig. 8.4. Strictly
speaking, you only need a preliminary support for the first element. When you have
8.2 Cupolas 141

Fig. 8.4 Building a modular

cupola

established the first full ring of plane trapezoidal elements and grouted the joints
between them, you can repeat the process with a second and a third ring. The result
is a cupola as seen in cross-section at Fig. 8.5. It consists of only three different
mass-produced plane trapezoidal concrete panels. It has a hole at the top that you
may cover with a panel without any structural importance or leave it as an opening
or a window.
It would often be beneficial to design the trapezoidal panels as cassette elements as
shown in Fig. 6.12. The outer frames of strong concrete then transfer the compression
forces along the medians (the longitude lines) and the lateral compression forces
along horizontal rings.

Fig. 8.5 Cross-section of a cupola of three elements

142 8 Shells

Fig. 8.6 Construction of element cupola [3]

You can make the cross-section of the frames more slender in the upper layers and
thereby reduce the weight and change the force distribution. Usually you would keep
the same depth of the frame profiles in order to have a joint between the elements
that can be grouted easily and be easily supported during erection.
Similar to the bottom blocks in the SL-decks, you may leave the light concrete
filling in the cassette elements without cladding and use it for acoustical damping.
A light aggregate concrete may give the same sound damping as a wood wool panel
that you often apply for this purpose alone. See more about this in Chap. 4. This
feature is very relevant, since a cupola acts as a concave acoustical mirror with an
unpleasant concentration of reflected sound.
The only tension reinforcement required is the pretension wires at the circular
foundation that you could establish as a circular pearl-chain of equal straight prefab-
ricated elements. The foundation elements could be produced with a length equal to
the bottom width of the above trapezoidal panels. The only scaffolding applied is a
pole at the centre or perhaps just a rod supporting a single element in each layer.
You may support the cupola by inclined arches as shown by Ottersland [3] in
Fig. 8.6. Four panels, each made as a pearl-chain arch, create a basis to support the
cupola.
You may post tension the upper circular edge of this basis, if the compression
obtained from the inclined arch structures is not sufficient for counteracting the
tension that the cupola above can give. See more about this in 8.2.3. You may coun-
teract the lateral forces from the arches by prestressing the floor slab in the lines
between the supporting points.
Calatayud [4] shows in Figs. 8.7 and 8.8 how to design an elegant cupola structure
of trapezoidal cassette elements. She designs it with a triangular floor plan and
combines two of them into a parallelogram floor plan, where inclined arches at the
edge panels allow a void for windows between the cupolas.
The structural principles presented here change the cupola from being an expen-
sive and time-consuming structure to be a fast, inexpensive and sustainable design
for large spaces.
8.2 Cupolas 143

Fig. 8.7 Triangular cupolas

[4]

Fig. 8.8 Triangular cassette element cupolas [4]

8.2.3 Cupola Design

When you calculate the forces in the cupola, you start from the top, where you have
no upper loads. Figure 8.9 shows the situation, when you reach layer number i of
height Hi with an upper horizontal radius ri and lower radius ri+1 .
Here you consider one of the trapezoidal elements with a load Pi at the horizontal
distance di from the bottom edge. You already know the accumulated vertical load P
on this element from the upper layers that will act at the distance ri+1 − ri from the
bottom edge.
144 8 Shells

Fig. 8.9 Force distribution in a cupola of trapezoidal panels

You now find the moment of these two loads about the bottom edge shown as
Point A in the middle of Fig. 8.9.

Mi = P(ri+1 −ri ) + Pi di

From that, you determine the horizontal forces Vi required to counterbalance this
moment and the force per meter at top and bottom. If you consider the top width
witop , you find this force per meter σitop from

Vi = Mi /Hi , σitop = Vi /witop

This force per meter represents a circumferential compression force

Fitop = ri σitop

It may be easier to comprehend this by considering the horizontal load as a

uniformly distributed “hydrostatic” pressure σitop in a slice of a cylinder as indi-
cated to the right at Fig. 8.9. This pressure acts on the wall of the cylinder, and it acts
in the cross-section of the cylinder in the figure. The force Fitop must therefore be in
equilibrium with the pressure over a length corresponding to the radius ri .
Now you do the same in order to find the circumferential tension force Fibot at the
bottom of the layer number i and the compressive force Fi+1top at the top of the layer
i + 1 below. The difference between these forces gives the resulting circumferential
compression force at the bottom of layer i as

Fres = Fi+1top − Fibot .

For flat cupolas, you will get compressive circumferential forces at all levels
except at the bottom, where you as mentioned have to prestress the ring foundation
to resist the tension force.
8.2 Cupolas 145

In case you design a cupola, you will get tensile forces in the shell, and you
can provide the bottom of the trapezoidal elements with ducts or grooves for post
tensioning a ring reinforcement at the layers, where it is relevant.

8.2.4 Cupola Element Design Example

Consider an element cupola with a diameter Dcupola = 30 m, and a hole in the top with
diameter Dhole = 5 m. The total height is hcupola = 10 m. The cupola has a shape as a
spherical cap. It consists of flat trapezoidal elements in three layers, with 18 elements
in each layer. All elements have the same length, but the width will decrease, as we
get closer to the top hole.
We wish to find the size of the elements.
Due to the spherical shape, any cross section through the centre of the cupola will
be circular with diameter r = 16.25 m, and the equation of the circle part is
  
y(x) = h cupola − r − r 2 − x 2

The arch length s is found by determining the cupola start angle v (Fig. 8.10)
 
Dcupola
v = arcsin = 67.4 ◦
2·r

Fig. 8.10 Cupola cross section

146 8 Shells

s = 2 · v · r = 38.2 m

The cross section arch length is subtracted with the arch length of the top hole,
and divided with the six concrete elements in the cross section to find the arch length
of each element selement = 5.53 m. Reusing the above equations, we can find the
straight element length Lelement when knowing the half angle between the radius r to
the ends of each element, νelement = selement /(2 r) = 9.76 deg:

L element = 2 · r · sin(velement ) = 5.51 m

To determine the width of the elements in each of the four levels of the cupola (A,
B, C and D), it is necessary to find the diameter of the spherical cap at each level.
For instance, the angle from vertical to the point B is νB = ν – 2νelement = 47.87 deg.
The x-coordinate of point B is then xB = r sin(νB ) = 12.05 m.
This is also the radius of the spherical cap at the level of the connection between
bottom- and middle element, and hence, the circumference length is

Ob = 2π · x B = 75.72 m

Now, the circumference length is divided into the 18 elements of equal widths sB
= OB /18 = 4.21 m, and the straight width wB (flat elements) is determined from the
start angle of the circle parts
 π 
wel B = 2 · x B · sin 2 · = 4.19 m
2 · 18
The same method can be used for the remaining levels, and the final widths of
elements in each level are:
 π

• Level A (cupola bottom): wel A = Dcupola · sin 2 · 2·18 = 5.21 m
• Level B: wel B = 4.19 m
• Level C: vc = v B − 2velement = 28.36 deg, xc = r · sin(vc ),
 π 
welC = 2 · xc · sin 2 · = 2.68 m
2 · 18
• Level D (edge of hole in top): v D = vC −2velement = 8.85 deg, x D = r ·sin(v D ),
 π 
wel D = 2 · x D · sin 2 · = 0.87 m
2 · 18
The flat elements can now be designed, for instance as cassette elements (similar to
Fig. 6.12), where there can be positioned post-tensioning cables around the circum-
ference in levels where it is required, to avoid cracking from the ring forces of
the cupola. The ring forces can be found as explained in 8.2.3, when the element
geometry and therefore the dead load is known.
8.2 Cupolas 147

8.2.5 Half Cupolas

As seen from Fig. 1.13, half cupolas have also been popular structures for thousands
of years.
The force distribution of a half cupola is more complicated than for a full.
However, a rather simple approach for direct engineering (see Sect. 3.2) may be
to design it as explained above for the full cupola and introduce an arch of strong
elements at the edge of the opening as a curved beam with lateral load.
The elements at the curved edge beam are subjected to shear and moments from the
horizontal forces. These edge beam elements therefore probably need to be massive
and made of strong reinforced concrete.
Half cupolas may have a number of applications. They may for example serve as
pleasant cool shelters in landscapes. They can also be used to accommodate theater
performances and concerts for an open-air audience by creating a shelter for the
artists and at the same time constitute an acoustical reflector.
By choosing the quality and type of materials for parts of the inner surface of the
half cupola, you can support sound reflection as a function of frequency so that you
for example can damp the deep tones and support reflection of the high tones, which
often are reduced the most on their way to the audience. See more about acoustical
damping in Sect. 4.1.5.

8.2.6 Torus Part Shells

You can design torus part shells by means of a technique similar to the one used for the
cupolas. Again, you may apply for example only three trapezoidal elements and again
cassette elements could be advantageous. Genius master builders like Apolodorus of
Damascus made torus part shells as seen from Fig. 1.11 where he adopted the lateral
forces as compression in a horizontal ring beam supported by columns to the left in
the photograph.
It is possible to combine a cupola and a torus part shell as shown in Fig. 8.11.
A vertical cylindrical wall or tube at the edge of the cupola supports both. The
horizontal forces from the torus part are resisted by a circumferential prestressed
cable at the outer edge. All other parts are subjected to compression, and you may
design the structure by the same principles as shown for the cupolas by means of
moment equilibrium about a lover edge of each element level.
As also indicated in Fig. 8.11 in the top drawing, you can apply a filling for
example with a light concrete, at the top of the torus and cupola, if you want to apply
the structure as floors in for example a round tower. You can supplement this with a
circular concrete slab that can assist carrying the load by resisting tension in the radial
direction. If the filling can transfer shear, the torus will act as a sandwich arch with
increased moment capacity. The slab can be prestressed by the compression provided
by tensioning the circumferential prestressing cable, so that tension will only unload
148 8 Shells

Fig. 8.11 Torus part shell and cupola supported by a cylindrical wall

the compression, and the large stiffness of the concrete is therefore utilized. A special
application for a structure like this is reutilization of silos.

8.3 Modular Shells

8.3.1 Modular Shell Structures

Above, we have dealt with vaults, cupolas and torus part shells. However, the shell
concept comprises infinitely many structures of which a good share can benefit from
the technologies of light and super-light concrete. Designers can make curved strong
spines of simple mass-produced concrete parts as load-bearing parts.
They can then fill out the space between the spines with light concrete cast against
double curved textile moulds. The moulds can be attached to the spines and perhaps
be made by inflated air or provided partly with inflatable air channels and stiffened
with rods in order to create the desired curvature.
These technologies already exist and require only few resources. They remove
the main practical hindrances for application of shells for in-situ cast structures as
well as for shell structures assembled from modular prefabricated shell elements.
8.3 Modular Shells 149

8.3.2 A Modular Shell System

Figures 8.1 and 8.12 show examples of shells made by a system of modular structural
elements. Peters et al. [1] called the system Polylithic. The system consists of only
one single shell element shown to the left at Fig. 8.13.
The basic element consists of curved pearl-chain spines along the three edges and
a pearl-chain arch at the middle. A textile mould is fixed to the spines on which the
light concrete membrane can be cast. If you require a large number of equal elements
for example for a structure as the one shown at Fig. 8.12, it could pay to produce a
more permanent mould.
You can create a permanent mould by first casting a prototype element as explained
above and then use that as a form for casting a final mould. You can make that for

Fig. 8.12 Modular shell structure by Peters et al. [1]

Fig. 8.13 Modular shell system. Peters et al. [1]

150 8 Shells

example as a light concrete block that you treat with a strong surface layer for example
of a fibre-reinforced mortar densified with waterglass.
You can assemble two basic elements to a folded arch as shown at the middle of
Fig. 8.13. Six basic elements can be assembled to a three point supported dome shell
as shown to the right of Fig. 8.13. By combining domes, arches, and single elements,
you may create shell structures covering a large area as shown at Fig. 8.12.
You could create simple connections between the basic elements for example by
bolting steel plates that also serve as anchorage plates for the prestressing wires of
the pearl-chain spines.
This allows you to create a slit between the elements that can serve to light the
interior.
You may for example create the slit by inserting a piece of concrete between the
shells to keep them apart. You can then connect the anchor plates in two neighbour
shells with a bolt through a hole in the piece.
If you stress the bolt, you can post tension the concrete piece so that it can transfer
tension as unloaded compression and serve as a prolongation of the pearl-chains in
the edges of the two neighbour shells.
This means that the pearl-chains of more shell elements can act as a continuous
spine in the assembled shell structure.
At the same time, the concrete piece can protect the bolt against fire.
The shells can then appear lighter and at the same time, they can become easier
to erect. Figure 8.14 gives an impression of how this may appear.
This is only one example out of an infinite number of possible modular shell
structure solutions.
Another way to let in natural light could simply be to substitute parts of the light
concrete with windows.

Fig. 8.14 Modular structure with slits between the elements. Peters et al. [1]
8.4 Hyperbolic Shells 151

8.4 Hyperbolic Shells

8.4.1 Hyperbolic Shells

Hyperbolic paraboloids also called parabolic saddle surfaces have been quite popular
for in-situ cast concrete shells. The reason is that this simple double curved surface
may provide a natural distribution of forces and that it contains two sets of straight
lines. This means, that you can create inexpensive moulds for casting shells of this
shape from straight wooden boards. Architects like for example Antonio Gaudi [6]
and Félix Candela [7] investigated the static possibilities of these shell structures and
showed how they can be applied for a great variety of architectural expressions.

8.4.2 Hyperbolic Element Shells

It is tempting to think that you can design hyperbolic shells from a number of equal
elements all with the same twist measured as increase of inclination per unit length.
Imeri [5] investigated the possibilities and soon realized that twisted elements as
those to the left in Fig. 8.15 would be of different shape although they all have the
same twist and they all have an equal quadratic projection on a horizontal plane.
She realized that she could create a saddle shell shown to the right of Fig. 8.15
from trapezoidal elements. However, most elements would be different, because the
arches across the saddle shape are parabolas. If you replace the parabolic arches with
circles, you may design an approximate hyperbolic saddle shell from only a limited
number of different trapezoidal elements.

Fig. 8.15 Hyperbolic shells [5]

152 8 Shells

8.5 Tubes

8.5.1 Tubular Structures

Tubular structures may have many purposes such as pipes for transport of fluid or
gas, sewers, tubes enclosing and protecting conveyors, tubes enclosing and carrying
footways, roads, subways, pneumatic post systems or even pneumatic train transport
systems.
An example is the tube bridge structure shown in Fig. 8.16. It is seen from the inside
in Fig. 8.2. Here, the tube not only serves to cover and protect an internal footway,
but also serves as a beam. The perforated sides of this post-tensioned pearl-chain
tube transfer shear between top and bottom of the beam, which makes it possible to
place it on pillars, so that this new element of infrastructure can be elevated above
an existing developed area. The shear in the tube also transfers torsion, so that the
tube can have a curvature in the horizontal plane between the supports.
Designers making tubes applied for transport have to fulfill a number of common
functional requirements. The tube should protect the interior against outer effects
from the climate or water pressure. Especially, the tube should be able to withstand
the effect of a tube fire. Tube fires are usually violent, because heat can only escape
in the longitudinal direction. This gives rise to jet fires with high temperatures and
high-speed propagation.

Fig. 8.16 Super-light tube bridge by Herdal et al. [2]

8.5 Tubes 153

Fig. 8.17 Drilled tunnel under Great Belt in Denmark

8.5.2 Tunnels

It is tempting to design a tubular tunnel of high-strength concrete, because it has a

strength to resist an outer mechanical load, and it is dense and therefore able to resist
an outer water pressure. However, high-strength concrete is susceptible to explosive
spalling and that in combination with powerful tunnel fires can give rise to critical
damages.

8.5.3 Drilled Tunnels

If the tunnel is circular, heating also gives rise to thermal compressive stress at the
inner surface, which further increases the risk of explosive spalling.
We have experienced this effect for example at fires in the Great Belt tunnel in
Denmark (Fig. 8.17) and the Channel tunnel between Britain and France. See more
about tunnel fires and explosive spalling in Hertz [8].
Tunnel walls made of high-strength concrete therefore require a protection against
fire by for example an insulating lining for example of mineral wool.
A super-light tunnel profile solves this problem.
By placing the high-strength concrete at the outer surface, and combine it with a
light concrete, the high-strength concrete can provide the denseness, and the inner
light concrete can protect and stabilize the strong.
A tunnel profile with this type of cross-section could be made of elements mounted
behind the head of the drilling machine in a continuous process.

8.5.4 Immersed Tunnels

A traditional immersed tunnel like the upper one in Fig. 8.18 has meter thick walls and
a heavy reinforcement to sustain the bending caused by the outside water pressure.
154 8 Shells

Fig. 8.18 Immersed tunnels: Traditional and super-light

The thick walls also provide the tunnel with a weight required to keep it in place
at the seabed.
A super-light immersed tunnel can for example have a cross-section as the bottom
one in Fig. 8.18. Here, the principle of the cross-section is the same as suggested for
the drilled super-light tunnel with a dense high-strength concrete at the perimeter and
a light concrete at the interior. However, the elliptical shape means that the vertical
hydrostatic pressure is larger than the horizontal. You can then provide the cross-
section with vertical interior walls for example made of ordinary concrete to sustain
the part of the vertical water pressure that is larger than the horizontal and design the
elliptic high-strength concrete to sustain a compression equal to the horizontal water
pressure.
This tunnel profile saves concrete, reinforcement, and CO2 in production. It is
therefore light and easier to tow. In order to immerse the profile to the seabed, you
can provide it with cavities, which you can fill with sand. Because the forces are all
in compression, you can save reinforcement.
In total, the super-light immersed tunnel requires considerably less concrete and
reinforcement, and it will therefore be more sustainable.

8.5.5 Floating Tunnels

Immersed tunnels are preferable, where the seabed is even and not too deep.
However, often you have conditions, where this is not the case.
For example, the Norwegian west highway should pass fjords with a depth of
1300 m, and here a floating tunnel may be desirable.
8.5 Tubes 155

Fig. 8.19 Floating super-light tunnel

You may also have a rocky seabed, where it is possible to immerse the tunnel at
certain parts, but where it will be very costly to provide an even foundation at other
parts.
Here a combined solution may be an alternative, where you apply the same profile
in the full length, but where it is floating in parts, where you cannot support it by an
even seabed.
The proposed immersed elliptic tunnel design from 8.5.4 is therefore prepared for
application as a floating tunnel (Fig. 8.19). In 8.5.4, we explained how some cavities
could be filled with sand. Here, we can adjust the degree of filling to balance a desired
weight of the tunnel, so that it can float and be stable for all load conditions.
Usually, it will pay to design a floating tunnel with a curvature in the horizontal
plane and provide it with a longitudinal post tensioning as a long pearl-chain.
This ensures that it can counteract a sideward pressure from streaming water.
When viewed from above, the entire tunnel then acts as an arch in compression,
or as a tension chain, depending on the direction of the water flow.

8.6 Floating Structures

8.6.1 Floating Structures

For floating bridges, it is also relevant to apply a curvature in the horizontal plane as
mentioned for floating tunnels in Sect. 8.5.5. Bridge elements can be post-tensioned
as a pearl-chain, resting on a series of caissons, and the entire bridge can hereby
serve as a horizontal compression arch or tension tie for lateral pressure of streaming
water.
An example is shown in Fig. 8.20.
Floating concrete structures are not something new.
156 8 Shells

Fig. 8.20 Floating bridge project. Drawing by Lund [9]

According to Pliny the Elder [14], Emperor Claudius ordered a high concrete ship
to be built at Pozzuoli, where there is plenty of fly ash as explained in Sect. 1.1.2. A
huge ship made for transport of Caligula’s obelisk at St. Peter’s Square was reused
as a permanent mould. The floating concrete building was sailed and immersed as a
part of the tower structure for the lighthouse at the harbour at Ostia. Archeologists
have now identified the place near the Fiumicino airport of Rome.
In modern times, the Norwegians have built huge floating concrete structures as
for example the Troll A platform. With a height of 472 m and weight of 684,000
tonnes, it was the tallest structure to be moved, when they towed it 200 km to the
Troll oil and gas field north-west of Bergen in 1995.
You can find many other examples of caissons of concrete. You can also find
numerous examples of houseboats made of concrete by specialized companies or as
self-made structures.
Usually, they make a floating concrete hull and place a house of low weight on
top.
A friend of one of the authors made such a houseboat, and called for advice,
because the hull was leaking. When the author came to the place after a week, the
concrete hull had re-sealed itself due to delayed hydration of the cement.
Experience show that such concrete structures in general are easy to maintain and
repair when compared to similar floating structures of steel and wood, which are
susceptible to rust and rot and need to be repainted.
Figure 8.21 shows a simple example of a floating building project by Jensen et al.
[10]. They used a number of super-light flat elements with high-strength concrete at
the outside and light concrete at the inside like the structure of the super-light floating
tunnel in Sect. 8.5.5. First they placed a number of these elements in the shape of a
capital G and joined them with a water-tight mortar. Then they post-tensioned them
together as a pearl-chain.
Then they aligned several G-shaped profiles and post-tensioned them together
across.
8.6 Floating Structures 157

Fig. 8.21 Floating restaurant building. Drawing by Jensen et al. [10]

This made a long two-storey building, where the basement could float as a hull.
Columns kept the three deck levels apart.

8.6.2 Sub-Water Cast Floating Structures

Figure 8.22 shows the principle of casting a caisson for the floating bridge shown in
Fig. 8.20 by Lund [9].

Fig. 8.22 Caisson for floating bridge. Drawing by Lund [9]

158 8 Shells

She placed an inflated ring-shaped balloon at the surface of the sea carrying a
submerged water filled permanent textile mould with the shape of a caisson.
She provided the mould with carbon fibres as reinforcement and with channels to
be cast out with light and strong concrete.
By casting the textile mould under water, she managed to counteract the mould
pressure to be an absolute minimum required to maintain the shape of the caisson.
This means that she subjected the permanent textile mould to a minimum of stress.
She could therefore design it using a minimum of materials and releasing a minimum
of CO2 .
After casting a caisson, the water inside is pumped out. It can then float, and the
ring-shaped balloon can be re-used for casting the next caissons.
The principle of combined application of textile mould for sub-water casting of
light concrete seems to be an extremely beneficial and sustainable solution that can
be applied for many structures at sea such as pier foundations, walls for fish farms,
and hulls for floating buildings and ships.

8.6.3 Concrete Ships

During the First World War, President Woodrow Wilson ordered a flotilla of 24
concrete merchant ships to compensate for the lack of steel [11]. Twelve of them
were built by different companies such as San Francisco Shipbuilding Company and
by the River Jacksonville Concrete Shipyard.
In 1942, the U.S. Maritime Commission for the same reason contracted
McCloskey and Company of Philadelphia to build twenty-four concrete ships [12].
Many other concrete ships have been made. For example, Pier Luigi Nervi
constructed a 400-tonne coaster and a private ketch of concrete [13].
It is obvious, that textile moulds and the principle of super-light concrete structures
with a dense high-strength concrete at the outer surface and a light concrete at the
interior as shown in Fig. 8.23, is beneficial for building of concrete ships.
With this technology, you can create the main structures light, which means that
you save material, money and CO2 . In addition, you can produce curved and double
curved shapes without application of expensive moulds.
Where previous larger concrete ships have suffered from being heavy at all parts
of the hull, the super-light technology means, that you can place inexpensive and
CO2 free ballast exactly where you want it to stabilize a light hull.

Fig. 8.23 A super-light ship

8.6 Floating Structures 159

These benefits mean that you can also make a super-light ship more stable,
maneuverable, and seaworthy than previous concrete ships.

References

1. Peters LS, Bekakos G, Mitrovgenis E (2020) Polylithic—a super-light concrete structure.

Super-light project 2020–07. Department of Civil Engineering, Technical University of
Denmark, 40p
2. Herdal RQ, Ali MM, Husarciková M, Kilic M, Calderón JD (2020) Space-time bridge. Super-
light project 2020–10. Department of Civil Engineering, Technical University of Denmark,
29p
3. Ottersland J (2018) Super-light spatial structures. MSc project. Department of Civil Engi-
neering, Technical University of Denmark. 120p.
4. Calatayud AD (2020) Design of a concrete construction to cover the Elsinore freeway. MSc
project. Department of Civil Engineering, Technical University of Denmark, 168p
5. Imeri V (2019) A super-light shell structure. MSc project. Department of Civil Engineering,
Technical University of Denmark, 146p
6. Lorenzi MG, Francaviglia M (2010) Art and mathematics in Antoi Gaudi’s architecture: “La
Sagrada Família.” J Appl Math 3(1):125–146
7. Alanís, EDA (2008) Félix Candela 1910–1997 The mastering of boundaries. Taschen, 96p
8. Hertz KD (2019) Design of fire-resistant concrete structures. ICE Publishing, Thomas Telford
Ltd., 254p. London. ISBN: 9780727764447
9. Lund KOA (2018) Floating concrete foundation with textile fiber bag. MSc project. Department
of Civil Engineering, Technical University of Denmark, 122p
10. Jensen D, Josefsen MT, Jensen MR, Homann MH, Marquardsen T, Chatti Y (2015) KDY—
Tuborg Havn. Super-light project 2015–02. Department of Civil Engineering, Technical
University of Denmark, 64p
11. Stilwell B (2020) The US Navy built 12 concrete ships for World War I. https://www.wearet
hemighty.com/articles/the-us-navy-built-12-concrete-ships-for-world-war-i
12. De Graaf J (2020) Concrete ships of World War I and II. https://www.thefactsite.com/concrete-
ships-facts/
13. Desideri P, Nervi PLjr, Positano G (1979) Pier Luigi Nervi—A cura di Paolo Desideri, Pier
Luigi Nervi jr, Giuseppe Positano. (In Italian) Zanichelli Editore, Bologna, 215p
14. Pliny the Elder (79) Naturalis Historia (natural history) Book 36, Clause 70. Dansih translation
by Jabob Isager. Forlæns Publishing Company 2018, 412p
Chapter 9
Structural Detailing

Abstract Some structural details of Super-light structures are explained. The theory
of reinforcement anchorage is provided, and design and full-scale tests of concrete
hinges for pearl-chain arches are shown. Furthermore, a potential for easy disas-
sembly and reuse of structural elements like the SL-decks by using a weak mortar
in joints is discussed.

9.1 Anchorage

9.1.1 Anchorage Basics

In principle, anchorage failure of a reinforcing bar in concrete can occur in two ways:
(1) Bond failure, where the bar is pulled out of a round hole, and
(2) Splitting failure, where a longitudinal crack develops along the bar to the
surface of the concrete structure or perhaps to other bars.
Bond strength is a property of the concrete and the corrugation of the bar.
Splitting strength depends on the geometry and strength of the cross-section.
The anchorage capacity is a minimum of bond- and splitting strength.
Abrams [1] made a pioneering work testing anchorage of a large number of
reinforcing bars with varying cover thicknesses in various concrete specimens.
Goto [2] showed that when you pull a corrugated reinforcing bar embedded in
concrete, shear will occur in the concrete next to the bar in planes through the bar
axis. Figure 9.1 shows to the left this shear τ at the bar surface.
Since shear is compression and tension, and since tension strength is smaller than
compression strength, cracks will radiate from the bar when the tension stresses
exceed the tension strength of the concrete. Goto showed by tests, that these conical
cracks radiate at 45° from the bar. Figure 9.1 also shows such cracks at the bar to the
left.
When the cracks are formed, you cannot transfer shear any more in the concrete
next to the bar. You may still be able to increase the anchored force if the structure can
react against it with inclined compression stresses as also shown to the left in Fig. 9.1.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 161
K. D. Hertz and P. Halding., Sustainable Light Concrete Structures, Springer Tracts
in Civil Engineering, https://doi.org/10.1007/978-3-030-80500-5_9
162 9 Structural Detailing

Fig. 9.1 Anchorage of reinforcement

This requires that the concrete can resist the inclined compression at the bar surface,
where this compression is largest. If not, you get a bond failure. It also requires that
the structure can resist the part τ of the inclined compression that radiates from the
bar. If not, you get a splitting failure.

9.1.2 Splitting Strength

Tepfers [3] showed in his thesis and later in papers like [4], that a required radiating
compression on a reinforcing bar may give rise to a longitudinal splitting crack along
the bar to the surface of the structure. The right side of Fig. 9.1 shows a cross-section
with such a longitudinal crack. The first author developed in part 2 of his thesis [5]
and in a later paper [6] a method for calculation of splitting strength.
If the shear stress at the surface of a bar of radius R is τ, the surrounding struc-
ture should resist a radial compression stress of the same magnitude τ, after the
conical cracks are developed. This radiating compression stress will give a hydrostatic
pressure of τ in the reinforcing bar.
Figure 9.2 shows that this hydrostatic pressure is equal to a compression force
2τR over a diameter of the anchored bar per unit length of the bar. Considering
equilibrium at a vertical cross-section through the centre line of the bar, you see that
the surrounding structure should resist a tension of τR at each side of the bar. This
means that the structure in the cover of thickness cs to the nearest surface should
resist a tension force of τR per unit length of the bar.
The right side of Fig. 9.1 shows a cross-section of a bar and forces in the cover
resisting formation of the splitting crack. A main contribution to this resistance is
the tensile strength of the concrete fct over the cover thickness cs .
We often apply a cross-reinforcement with area Ats per unit length in order to resist
this force and codes of practise often prescribe the use of it. However, if we utilize
the full tensile strength of that, we cannot take the tensile strength of the concrete
into account, because the concrete cracks long before the reinforcement reaches its
9.1 Anchorage 163

Fig. 9.2 Hydrostatic

stresses in a bar

strength. In most cases, the tensile strength of the concrete is larger than the tensile
strength of the cross-reinforcement. This means that, if we want to consider the cross-
reinforcement in addition to the concrete tensile strength, it can only contribute with
a tensile stress of σts = 31.5 MPa, which it will have at a strain of εs = 150·10–6 ,
where the concrete cracks. (σts = εs Es ).
Very often, we would like to calculate the splitting capacity at a bearing for
example at the end of a beam or slab element. In such a zone, we may have a vertical
bearing stress σb from the reaction force. The reaction usually gives a transverse
compression equal to the Poisson ratio ν of the concrete times the bearing stress σb .
We can add this compression νσb to the tensile strength fct of the concrete, because
the splitting force should first counteract the compression before it can use the tensile
strength.
Above, we have seen that the tensile strength of the structure in the cover thickness
should resist τR, where τ is the shear at the surface of the anchored bar. We therefore
find the anchorage force per unit length of the bar as

2π Rτ = 2π (Ats σts + cs (fct + νσb ))

The effect 2πcs νσb of the bearing stress σb explains why it is possible to apply
anchorage lengths of only 55–70 mm at the end of many prefabricated concrete slabs.
These small anchorage lengths are required if slab elements from two sides should
rest on the same wall with a thickness of 150–200 mm. Element factories have previ-
ously shown this by full-scale tests, because the empiric rules in different textbooks
and codes of practise would typically require 10 times longer anchorage lengths.
164 9 Structural Detailing

The method for anchorage calculation presented above therefore represents a

major improvement for assessing the anchorage capacity and for optimal design of
the material consuming bearing zones of super-light elements.

9.1.3 Bond Strength

After formation of conical shear cracks as shown to the left in Fig. 9.1, the structure
may be able to resist anchorage shear transfer by inclined compression stresses.
These compression stresses will be largest near the surface of the bar. Bond failure
happens at a pullout force, where these inclined compression stresses exceed the
concrete strength and the anchorage force pulls the bar out of a round hole.
In Sect. 3.2.2, we found that the distribution of forces and stresses at an ultimate
limit state gives a maximum resistance to load and deformation.
If the angle between the bar axis and the direction of the ultimate inclined concrete
stress σcau is called α, the transferred shear stress τ at the bar surface will be

σcau sin 2α
τ = σcau sin α cos α =
2
From that, we see that a structure obtains the largest shear capacity and pullout
strength for a given strength of the concrete, if the compressive stresses are inclined
45° degrees to the bar axis as also presumed in Sect. 9.1.2.
If you recall that the inclined compression in the concrete is conical around the
bar, you may see that the ultimate compressive stresses in the concrete must the
ultimate strength of the concrete with hindered lateral expansion.
Because for each radiating lamella, where you consider the inclined compres-
sion, there will be neighbour lamellas with the same compression, so that sideward
expansion is not possible.
Since the biaxial concrete strength with hindered lateral expansion is about 1.3
times the monoaxial compression strength fcc as shown to the right of Fig. 9.3
according to [7, 8], you may presume that σcau = 1.3 fcc and

1.3 fcc sin(2*45)

τ= = 0.65 fcc
2
provided that the corrugations on the bar surface can transfer the full ultimate
shear to the concrete.
In order to investigate the shear transfer from bar to concrete, the Nordtest foun-
dation supported the first author in making a test method called the cuff test shown to
the left in Fig. 9.3. A reinforcing bar is cast in the centre line of a concrete specimen
shaped as a half standard cylinder prolonged in a 45° cone.
9.1 Anchorage 165

Fig. 9.3 Cuff test for bond strength and biaxial strength of concrete

The specimen is placed in a steel bearing block—the cuff—with a conical cavity

provided with a layer of artificial rubber (neoprene) to distribute an effect of small
irregularities.
At first, the test series comprised 280 specimens with different shapes of the
reinforcing bars and diameters varying from 8 to 25 mm exposed to temperatures
from 20 to 400 °C.
The test series soon proved, that an ordinary deformed reinforcing bar as the one
shown in Fig. 9.1 has a bond strength of τ = 0.65 fcc as estimated theoretically above
to be the maximum possible [5, 6]. The test series also proved the validity of the
expression, when the concrete quality varies for example according to an increased
temperature.
For plain round bars, you cannot obtain the same bond. The deformations at
the surface of these bars are so small that they can lose their bond by an elastic
deformation of the concrete. This has often been demonstrated at lectures by knocking
the bar of the specimen on a desk.
The concrete block loses its bond by the chock and moves typically 150 mm down,
where it again grips at the bar. You can repeat this up and down about 50 times before
the inner hole in the concrete becomes so worn that the elastic bond is permanently
lost.
The elastic bond strength of round bars were shown to be at least τ = 0.15 fcc .
For some bars, rust may increase this value, but it is not recommended to rely on
more.
Later, a new study was made, where the bond was investigated of twisted wires
applied for prestressing lines [9]. Here we found a safe value of the bond to be τ =
0.25 fcc .
The test series showed that the same expression was applicable for varying
diameters of the wires and for varying concrete qualities.
166 9 Structural Detailing

The good agreement between test results and theoretical obtained values meant
that the cuff test was not needed for anything else than documenting the formulas.
The cuff test has therefore not been introduced as a possible new ISO standard test
method.

9.1.4 Anchorage Summary

You can estimate the anchorage capacity as a force Fa per unit length of a reinforcing
bar of diameter D with a cover thickness cs from the centre line to the nearest surface
in a concrete structure.
Fa is the smallest of the splitting capacity Fs and the bond capacity Fb .
Consider a concrete with compressive strength fcc .
The tensile strength fct is typically about 0.1·fcc .
The Poisson ratio ν is typically between 0.1 and 0.2.
In a reaction zone, you may have a stress σb across the cover thickness.
You may have a cross-reinforcement of area Ats per unit length.
From that, you get a splitting capacity per unit length of the bar of

Fs = 2π (Ats 31.5 MPa + cs (fct + νσb ))

and a bond capacity of

Fb = π D k fcc

The factor k depends on the shape of the bar:

• For deformed bars, you can obtain the maximum theoretical value k = 0.65.
• Tor prestressing lines and other wires k = 0.25.
• For round bars k = 0.15.
You can find more information about anchorage at high temperatures for fire safety
design in [10].

9.2 Separable Joints

9.2.1 Lime Mortar Joints

For thousands of years, masons have applied hard-burnt bricks joined with lime
mortar for walls, aches, and vaults. The lime mortar transferred compression and
served to glue the bricks into a composite material. It ensured dense, heavy, and
sound insulating structures. The relatively weak lime mortar also allowed builders to
9.2 Separable Joints 167

pull down old structures for example by knocking the walls sideways. After that, the
hard-burnt bricks were mostly intact, and it was possible to clean them by knocking
residual lime mortar off.
Masons could then re-use the bricks in new structures. Builders could apply the
debris of lime mortar as a filling material in roads or buildings, or as aggregate in
concrete structures. The debris could also be grinded into gravel and applied in new
mortar.
When masons introduced cement-based mortar, brickwork became stronger, but
also more difficult to pull down and to clean and re-use the bricks.
From this experience, we can learn deliberately to apply the weakest possible lime
mortar for joints in order to open an easy possibility for reusing the joined elements.
We usually construct super-light elements with a strong and heavy concrete of
typically 55 MPa at the bearings of the elements to make them robust. Here we also
introduce forces and anchor possible reinforcement, and in addition, the weight of
the heavy concrete does not contribute much to the moment distribution. The builder
can then design the whole structure so that the mortar of the joints should as far as
possible only transfer distributed compression stresses. Lime mortar is suitable for
this purpose. The weak lime mortar with small stiffness will in addition contribute
to distribution of bearing stresses.
A good example of a design, which is suitable for lime mortar joints, is the
beamless super-light decks in Sect. 4.3. Here, we have designed the blade connection
in Fig. 4.22 to transfer reaction by distributed compression stresses in the mortar.
We have also deliberately placed the connections where we do not need transfer
of moments as illustrated in Fig. 4.21. The beamless super-light decks are also an
example of a system of elements of standard dimensions. We can apply this system
in many buildings and it has therefore a potential for being dismantled and re-used
[11].
As discussed in more detail in Sect. 11.1.1 and shown in Fig. 11.6, you cannot
decide that your new component should be re-used at end of life. However, you can
decide to re-use a previously designed component and harvest the benefits of a low
resource consume and CO2 emission or make your new component more attractive by
making it re-usable. If your structure was part of a building system with standardized
dimensions and joints such as the beamless super-light decks, it becomes more likely
that it can be reused and more likely that you can find a suitable used component for
your structure.

9.3 Hinges and Weak Zones

9.3.1 Concrete Hinges

Hinges are applied in many structures to control the boundary conditions of the
static system. Usually, the assumption is that the construction can rotate freely and
168 9 Structural Detailing

without any resistance at the location of the hinge, and this may be a good conservative
approach in many cases. However, in reality, there exist many types of hinge designs,
where you must expect a rotational resistance.
For a pearl-chain arch or vault, a concrete hinge is an obvious choice, see Fig. 9.4.
Concrete hinges were widely used in arch bridges from the beginning of the twentieth
century. In the last part of the nineteenth century, Jules Dupuit was a pioneer in the
use of different types of hinges to control the thrust line of masonry arches [12].
However, it was not until 1907 that the French engineer Augustin Mesnager thought
of designing a hinge made of concrete [13]. The Mesnager hinge was made for
concrete arch structures. The concept was simply to reduce the cross section height
of the arch structure at the zone of the hinge. That zone is referred to as “the throat”.
Mesnager applied crossing steel bars in the throat to ensure the integrity. The famous
builder Eugene Freyssinet read the work of Mesnager and developed the concrete
hinge further for use in arches. He implemented it into many of his large constructions
[14].
Freyssinet found that the crossing bars in the throat were not necessary when used
in arches, despite of the reduced cross sectional height. The reason was that arch
structures have a significant normal force through the hinge, and where the normal
stresses approach the throat zone, they will bend in a bottleneck shape. This direction
change causes a confinement and two-dimensional compression of the concrete in the
throat, which increases the capacity. Nevertheless, Freyssinet did apply the crossing
steel bars in his building to be on the safe side. Furthermore, the direction change
of the normal stresses causes a splitting force in the zone next to the throat, and this
would later be the focus of several studies by other researchers.
As Freyssinet began to apply concrete hinges, other engineers and builders would
follow, and many great arch bridges from that time utilize the concrete hinge. Some
well-known examples by Robert Maillard can still be found today in Switzerland
[15, 16].

Fig. 9.4 Section cut of

typical concrete hinge with
crossing steel bars in the
throat
9.3 Hinges and Weak Zones 169

Nowadays, concrete hinges are mostly applied in bridge piers [17], but the reintro-
duction of arches by using the pearl-chain principle, may also make a reintroduction
of concrete hinges relevant.

9.3.2 Tests of Concrete Hinges

Several full-scale tests have been performed at DTU on concrete hinge specimens
to verify their application in pearl-chain arches. All tests were to fracture. Tests
specimens with different hinge throat sizes were tested with a large level of normal
force corresponding to a low 30 m fully loaded pearl-chain bridge. Such low arch
would have a very high level of normal force, which was the reason for testing. The
test setup was a specially developed four-point bending with a constantly applied
normal force. The results of the response of hinges with crossing steel bars is shown
in Fig. 9.5 for three ratios between throat height b1 , and full cross section height b2 .
The hinge with the biggest throat of b1 /b2 = ½ was not promising for application in
pearl-chain arches due to cracking outside the throat. More details can be found in a
paper by the authors [18]
The present guidelines for the geometry of the hinges and their reinforcements are
originating from research by Leonhard and Reimann [19]. They showed that it was
possible to create a universal concrete hinge response curve, by which the behaviour
of any properly designed hinge could be anticipated.
Their analytical findings are verified by their own tests and tests by others [20–22].
The basic assumption of the universal hinge response theory is that there is
cracking in the hinge throat, and that it is possible to relate the crack length to the
level of normal force and moment in the hinge. The horizontal axis of the universal
hinge curve is the rotation α divided by a parameter K expressing the strain distribu-
tion, and the vertical axis is the parameter m expressing moment related to normal
force and throat height

Fig. 9.5 Results of hinge

responses to loading for
different geometries
170 9 Structural Detailing

α 1
=
K (1 − 2m)2
8·N
K =
9 · b1 · w · E 0
M
m=
N · b1

E0 is the stiffness of the concrete in the throat and w is the width of the throat
area.
Leonhardt and Reimann determined the above equations, and hereby, they could
set some limits for when the hinge throat would start to crack (m = 1/6), and when
the crack had reached half the throat (m = 1/3). Furthermore, they argued that a
hinge should never be designed to carry load beyond the m = 1/3 limit.
The test results from our tests of concrete hinges were compared to the universal
curve from Leonhardt and Reimann, and they showed that hinges with a large normal
force would still fit to the universal hinge curve, see Fig. 9.6. Later, hinges with carbon
fibre bars crossing in the throat were tested in full-scale as well. Results of those tests
are also shown in the figure. The benefit of using carbon fibre bars instead of steel
is that they cannot corrode, and hence requirements to concrete cover layers can be
ignored. The CFRP-bar hinge tests gave result with a similar close fit to the universal
curve [23].
It is recommended to apply hinges with the ratio b1 /b2 = 1/3 in order to be sure that
the crack will develop in the throat. You may then calculate the moment distribution
in the structure and the rotation α of the hinge, by initially considering the hinge
ideal with no moment as a conservative assumption.
Then you may calculate the parameter m relating the moment M to the normal
force N and the throat height. You can then check that m is within the recommended
limits of 1/3 for the ultimate limit state ULS and 1/6 for the serviceability limit state
SLS.

Fig. 9.6 Test results fit to

universal hinge response
curve
9.3 Hinges and Weak Zones 171

Fig. 9.7 Position and section of end-element in pearl-chain

When using concrete hinges in the supports of pearl-chain arches, you must
consider whether the post-tensioning cable should be anchored before or after the
hinge. If the cable runs through the hinge throat, then the post-tensioning force will
add to the total normal force in the hinge and reduce m.
An alternative to the Mesnager hinge is to apply a specially designed support-
element, as the end-element in the pearl-chain. Such element was used in full-scale
tests to fracture of pearl-chain arches, see Fig. 9.7. The purpose of such end-element
is to create a space to anchor the post-tensioning cable, and spread the local force
to the whole cross-section. The transition from arch to foundation is similar to an
unreinforced concrete saddle bearing. This is possible, since it will always be in
compression.
The structure can rotate in the unreinforced connection between arch and foun-
dation, if a mortar is poured in between the two when erecting the arch [24]. The
stiffness and thickness of the mortar has an influence on the rotational resistance of
the joint. During the full-scale tests, it was observed that cracks would initiate in the
mortar layer between arch and foundation, and rotation would occur [25].

9.3.3 Weak Zones Guiding Forces and Locating Fracture

When concrete hinges are applied, you can assume that you have created a weak
spot in the construction. If you use a Mesnager hinge with b1 /b2 = 1/3, which is
the common geometry, the bending stiffness will be 27 times smaller than that of
the adjacent structure. Depending on the structure, you can therefore anticipate that
cracking will initiate in the concrete hinge throat first.
This means that you, e.g. in relation to arch structures hinged at the supports, can
quickly assess the status of the arch by looking for cracks at the hinges. Furthermore,
the arch thrust line is guided through a hinge, and this may help you to control where
the forces are.
Since concrete hinges are not ideal, the structural response will be in between
the behaviour of ideal hinges, and that of fixed connections. Figure 9.8 shows an
172 9 Structural Detailing

Fig. 9.8 Moment distribution of loaded arch with and without concrete hinges

example of an arch with two hinges at the supports compared to an arch with fixed
ends.
You may choose to create an arch without hinges, but hereby you may not know the
location of where the structure will start to crack. The structure must then form four
plastic hinges before failure, unless the loading will cause a direct compressive failure
in the arch cross-section. Furthermore, the very rigid fixed arch will be susceptible
to settlements of the foundations, temperature variations, creep etc.
If you choose a structure with two hinges, one at each support, you have provided
the arch with the ability to withstand vertical settlements of the foundation. However,
horizontal settlements of the foundation may still cause some increase in the bending
moment. Still the arch is also susceptible to creep and temperature variations.
A three-hinged arch, with an additional hinge at the top, is statically determinate,
and therefore easy to approach as an engineer. Furthermore, it is not susceptible to
creep, temperature variations or settlements. Unfortunately, the three-hinged static
system will not provide any “reserve strength” if the foundation moves outwards due
to the arch load, and therefore this type of system should only be applied where the
soil conditions are good or you apply tension ties as explained in Sects. 6.2.5, 6.2.6,
and 7.2.3.
The authors full-scale tested two adjacent pearl-chain arches. Each was supported
by two saddle bearing hinges as illustrated in Fig. 9.7. The test showed a number a
warning signs before failure. The arches were loaded with a critical load around the
¼ point of the span. First, the cracking at the hinges initiated and propagated. Then
significant deformations were visible for the arch structure, and finally, two plastic
hinges formed and created a mechanism at failure load.
9.3 Hinges and Weak Zones 173

9.3.4 Super-Light Concrete Hinges

As explained in Chap. 3, the super-light principle allows the engineer to guide the
forces, where they are most useful by application of concrete of different density and
stiffness.
You can benefit from this by designing a concrete hinge as shown in Fig. 9.4, but
filling out the cavities on both sides of the reduced cross-section with a light concrete
as for example a foam concrete of density 300 kg/m3 (see Sect. 2.2.2).
In practise, you can cast these inserts first, and apply them as permanent parts of
the mould.
Doing so, you can make the mould for the hinge cheaper, and you obtain an
insulation of the hinge, protecting the concrete as well as any reinforcement in the
hinge against fire.
In addition, you can hinder stiff objects such as bottles etc. to fall into the groove,
and you obtain an even surface of the hinge, which again can serve for example as a
mould for casting a filling material in a sandwich arch as explained in Sect. 7.3.

References

1. Abrams DA (1913) Test of bond between concrete and steel. Urbana, University of Illinois.
Eng Exp Stat Bull 71:240
2. Goto Y (1971) Cracks formed in concrete near a reinforcing bar due to bond and transverse
cracking. J Am Concr Inst Proc 68(4):244–251
3. Tepfers R (1973) A theory of bond applied to overlapped tensile reinforcement splices for
deformed bars. Thesis. Publication 73:2, Division of Concrete Structures, Chalmers University
of Technology, Gothenburg, 328p
4. Tepfers R (1979) Cracking of concrete cover along anchored deformed reinforcing bars. Mag
Concr Res 31(106):3–12
5. Hertz KD (1980) Armeringsståls forankring ved høje temperaturer. (Anchorage of reinforcing
steel at high temperatures) in Danish. Report 138 and part 2 of Ph.D. thesis on fire exposed
concrete structures, Institute of Building Design, Technical University of Denmark, 103 p
6. Hertz KD (1982) The anchorage capacity of reinforcing bars at normal and high temperatures.
Mag Concr Res 34(121):213–220
7. Kupfer H, Hilsdorf HK, Rusch H (1969) Behavior of concrete under biaxial stresses. ACI J
66:656–666
8. Neville AM (1977) Properties of concrete. The Pitman Press, Bath, p 687p
9. Hertz KD (2005) Vedhæftningsstyrken af spændliner ved brandpåvirkning (Bond strength
of prestressing lines exposed to fire) (In Danish). Report SR 05–12 Department of Civil
Engineering, Technical University of Denmark, 10p
10. Hertz KD (2019) Design of fire-resistant concrete structures. ICE Publishing, Thomas Telford
Ltd, p 254p
11. Halding PS, Hertz KD (2020) Design for disassembly of super-light structures. RILEM spring
convention 2020
12. Holzer SM, Veihelmann K (2015) Hinges in historic concrete and masonry arches. Proc ICE
Eng Hist Heritage 168(2):54–63. https://doi.org/10.1680/ehah.14.00019
13. Mesnager A (1907) Experiences sur une semi-articulation pour voutes en Béton armé. Annales
de Ponts de Chaussees 2:180–201
174 9 Structural Detailing

14. Fernandez-Ordonez D (2018) Eugene freyssinet: I was born a builder. 28. Dresdner Brücken-
baussymposium, Technische Universität Dresden
15. Billington D (1984) Building bridges: perspectives on recent engineering. Ann N Y Acad Sci
424:309–324
16. Billington D (2000) The revolutionary bridges. Sci Am 283(1)
17. Morgenthal G, Olney P (2015) Concrete hinges and integral bridge piers. J Bridge Eng 21(1).
https://doi.org/10.1061/(ASCE)BE.1943-5592.0000783
18. Halding PS, Hertz KD, Schmidt JW (2014) Concrete hinges. In: Proceedings of the Iass-slte
2014 symposium
19. Leonhardt F, Reimann H (1965) Betongelenke—Versuchsbericht, Vorschlaege zur Bemessung
und konstruktiven Ausbildung. Deutscher Ausschuss Fuer Stahlbeton, pp 1–34
20. Base GD (1965) Tests on four prototype reinforced concrete hinges. Cement and concrete
association—research report, pp 1–28
21. Marx S, Schacht G (2010) Concrete hinges—historical development and contemporary use.
In: Proceedings of the 3rd international fib congress and exhibition, USA
22. Tourasse M (1961) Essais sur articulation Freyssinet (Experiments on Freyssinet hinges). Ann
Institute Technique du Batiment et des Travaux Publics 40(57):62–87 (in French)
23. Halding PS, Schmidt JW, Musachs S (2021). Static behaviour of concrete hinges with CFRP
bars. Future article in pipeline
24. Halding PS, Hertz KD, Viebæk NE, Kennedy A (2015) Assembly and lifting of pearl-chain
arches. Proc Fib Symp 2015:185–186
25. Halding PS, Hertz KD, Schmidt JW, Kennedy BJ (2016) Full-scale load tests of pearl-chain
arches. Eng Struct 131:101–114
Chapter 10
Sustainability

Abstract The combination of light aggregate concrete and ordinary concrete

decreases the environmental impact from super-light structures. The chapter provides
an overview of CO2 -emissions from different materials and building parts to compare
with super-light structures. The comparisons are made realistic by complying with
actual requirements to sound insulation.

10.1 Environmental Impact

The Environmental impact of our human activities has become more and more in
focus. The emission of CO2 and other greenhouse gases causes global warming
and climate changes, which have become visible as extreme weather. Temperatures
increase, water levels increase, the number of hurricanes, forest fires, and floodings
increases.
A climate debate is taking place, where people require solutions from politi-
cians and technicians. It is especially required that the present generation should not
transfer their CO2 problems to the next, but solve them instead (Fig. 10.1).
An early example of a study on these matters was a Ph.D. project by Andersen [1,
2], which was an answer to a question, only few had asked at the time. He found a
level for energy consume of 500–1000 kWh per m2 for dwellings equal to 150–300 kg
CO2 per m2 , which corresponds well to modern calculations.
This work combined with knowledge about ancient building technology inspired
the first author to develop new methods for constructing light sustainable concrete
structures.
He invented Super-light structures in 2008, and Pearl-chain structures in 2009.
In 2010, he and the Technical University of Denmark formed a start-up company
called Abeo Ltd. with a group of three students from Copenhagen Business School.
The name means, “abandon the old stuff” in Latin to mark a new beginning in
structural design.
The new technologies made it possible and economically feasible to produce
structures with less material consume and the same load-bearing capacities, improved
sound-insulating properties, and a better fire-resistance than traditional structures.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 175
K. D. Hertz and P. Halding., Sustainable Light Concrete Structures, Springer Tracts
in Civil Engineering, https://doi.org/10.1007/978-3-030-80500-5_10
176 10 Sustainability

Fig. 10.1 Masai Mara Kenya. Photo KD Hertz

In cooperation with architects, consulting engineers, and producers, the first

author, his students and the Abeo Company developed the first suggestions for super-
light alternatives to actual building projects. They calculated the CO2 emission for
structural parts and building processes for the lifetime of a number of projects. These
calculations demonstrated CO2 savings between 20 and 50% compared to ordinary
concrete structures, 80% compared to steel structures and up to 50% compared
to timber structures, when the calculations compared the ratio of CO2 release and
lifetime of the buildings.
In addition, the smaller consume of materials means less pollution from production
of materials and components and from erection of buildings. Application of clean
non-toxic materials such as lime, clay, sand, and water and new possibilities using
less carbon for heating them in the production processes means a further reduction
of pollution.
The new structures are therefore more climate-friendly and sustainable than tradi-
tional building technologies, and this is why the start-up company won the price as
“Best Danish Early-stage Cleantech Company” in 2010.
The same year it won the “Clean Tech Open Global Ideas Competition” in San
Francisco (Fig. 10.2). Byrne [3] declared in The New York Times:
Abeo thinks BIG for its Super-light Structures. For a company which has only been in
existence since June this year, Danish start-up Abeo has been racking up the awards for
rethinking an everyday construction material—concrete.

The word BIG was a reference to the Bjarke Ingels Group architects, which was
one of the cooperating companies that had just opened an office in New York City.
In 2013, the EU-commission gave Abeo Ltd. a prize as “Europe’s Best Innovative
Spin-out Company” (Fig. 10.3).
10.1 Environmental Impact 177

Fig. 10.2 Trophy for the world championship in clean technology won by the first author’s start-up
company in 2010. Photo KD Hertz

Fig. 10.3 Alexander Wulff from Abeo Ltd receives the prize as Europe’s Best Innovative Start-up
Company in 2013. Photo EU-Commission
178 10 Sustainability

However, in 2010 neither building industry nor politicians were willing to invest
in clean technology and CO2 savings. Abeo had to close down in 2016, and the
producer took over the company that today sells super-light deck elements in a
number of countries.

10.1.1 Systematic CO2 Assessment

You can often find many different values for the CO2 release of a product or a
material. Some producers find it tempting to give the assessment of their own product
a beneficial treatment, and sometimes they forget important details and contributions
to the calculation. Furthermore, you cannot apply fixed values for all products and
materials. Very often, you find considerable variations in manufacturing of a product
or a material from producer to producer depending on the processes applied and on
transport of raw materials to the location of production. You can also find considerable
variations for the same product or material made by the same producer at different
times. The producer may for example implement changes to make the product more
sustainable and of course, you should consider this in the assessment. Else, the
development has been in vain.
This means that you cannot make a fixed table with reliable definitive values
of CO2 for building materials and components. Instead, in each case, you have to
consider the processes involved for the particular producer at the time of production,
and even then, data are scattered due to random variations in the processes and due to
variation in the nature of the raw materials. As shown in Fig. 10.4, each process gives
a product and for that, you summarize the CO2 footprints. They consist of one part
for the raw material plus one part for energy such as heat, electricity or mechanics
plus one for human operation plus one of removing and treating waste products.
If the process gives a by-product, you may consider, if someone can apply it. In
most cases, the by-product is considered to be CO2 free by the receiver, and it should

Fig. 10.4 Process

10.1 Environmental Impact 179

not be subtracted from the account of the process. For example, a contribution to
district heating of a city is valuable with respect to CO2 , when it replaces heat from
a coal driven plant in wintertime. However, you cannot take this CO2 saving into
account, if the extra heat cannot be applied by the city in summertime, or if the plant
already has a fossil free energy production.
In this book, we are concerned about the impact of a building on the climate.
The CO2 emission damaging the climate comprises that of production of materials,
the building process, and the impact from the end of life stage. The latter includes
demolition of the building and processing materials as waste.
The end of life stage is especially important, because without that you may think
that you have a low CO2 impact, where you in reality only obtain to store your old
CO2 for a while.
Later, it is released damaging the climate for the next generations in addition to
their own CO2 contributions.
References like the standard EN 15804 and Herrmann [4, 5] therefore explain
this in detail. They emphasize the importance of the end of life stage and group
the processes involved in main classes A, B, and C. Each class is subdivided as for
example A is divided into A1-A5 (Fig. 10.5).
It is obvious that you should relate the CO2 emission of the life cycle including
the end of life phase to the lifetime of a structure or material. It is also obvious
that a comparison of climate impact of different building structures, components, or
materials can only make sense if they fulfill the same functional requirements for the
actual application.
For example, you can only compare emission of a carbon fibre reinforcing bar
with that of a steel bar, if they both have the same strength, and they are placed in
the structures so that the steel bar will not be damaged by rust and the carbon fibre
bar will not be damaged by fire exposure.
Some calculations include a benefit when combustible materials are burned at the
end of life stage and replace fossil fuels like coal in power plants producing district
heating. Likewise, oven heat from production of cement or light aggregates may

Fig. 10.5 Life Cycle phases according to CEN

180 10 Sustainability

contribute to district heating. This is only relevant, if the power plants are based on
fossil fuels, and if you can foresee that, they still will be that, at the time for end of
life perhaps 50 years from now. In most industrialized countries, this is not the case,
and such contributions are not included in the numbers presented in this book.
A potential for reuse will prolong the lifetime of a material or component and
reduce the CO2 impact related to the lifetime. The standard [4] suggests adding this
as a phase D. You should not consider this on beforehand, when you apply a material
or component for the first time in a structure and calculate the related CO2 impact.
At this early stage, you do not know whether an industry will reuse them at the
end of life of your building structure 50 years or more after construction.
Furthermore, a future building industry will be less willing to reuse your old
material or component, if you have already taken the benefit on your CO2 account.
If you for example apply a component that in a life cycle of phase A-C gives an
impact of 500 kg CO2 for a lifetime of 50 years, you have to report an impact of
10 kg CO2 per year.
If it were accepted that you instead presumed that someone would reuse your
component in a future building for another 50 years, you would only report the half
impact of 5 kg CO2 per year. In such a system, the future builder should also report
5 kg CO2 per year as you have done for a new component.
This means that the future builder has no incentive to reuse your old component
and no payment for the extra work and logistic of finding it and fitting it into the new
building.
We therefore recommend implementing a possible reuse phase D at the beginning
of the life cycle (Fig. 10.6). Here, you can be sure about how much you intend to

Fig. 10.6 Lifetime impact and absorption

10.1 Environmental Impact 181

reuse in your own construction, and you will have a clear benefit from doing so.
Recycled steel is a good example of an implemented technology that gives rise
to a considerably reduced CO2 impact that industry already takes into account as a
phase D at the beginning of the life cycle and not at the end of it.
The Ökobaudat database refers to the GaBi database [6] for actual CO2 numbers.
Other databases and sources of information like [7] from the University of Bath and
[8] from VTT, give values for the CO2 impact, which are only slightly different, when
you transfer their properties to numbers with the same meaning. This represents no
serious problem since the preconditions such as the nature of the sub processes, the
efficiency, and the time and place for deriving the numbers may be different.
The problem becomes serious, if you apply the values without including a possi-
bility for adjusting them according to well-documented improvements or variations
of the processes in question. If you do not do that, companies have no incitement for
changing their way of doing things, and your calculation method will be a hindrance
for development.
This means that you can apply general data from the databases assessing the CO2
impact. However, as soon as you have more precise and well-documented data for
an actual product, you should apply them instead.
In the following, we present estimates of the CO2 emission from (phase A) produc-
tion, (phase B) use, and (phase C) demolition of materials and products based on
accessible knowledge.
The values presented include equivalents for other greenhouse gases such as
methane and perfluorocarbons (PFC’s) etc.
For comparison, we also present ideal values for how little CO2 emission we can
reach, if all sub processes become as CO2 neutral as possible.
This indicates that we can expect considerable changes of the impact in the future.
It also shows how important it is that the basis for assessment of CO2 impact can
follow the changes, so that a too conservative assessment method does not hinder
the positive development (Fig. 10.7).

10.2 CO2 from Materials

10.2.1 Concrete

The building industry accounts for a large part of the global CO2 emission, and
cement production alone causes about 6% of the emission of the Earth. This amount
is increasing while building industry is developing mainly in China, but also in
India, and Africa [9, 10]. We therefore have to address technologies like those of this
book, by means of which, we can reduce cement consume and create structures with
application of less concrete.
However, when we make comparisons between alternative building materials and
methods in order to choose optimal solutions, it is important to be able to estimate the
182 10 Sustainability

Fig. 10.7 Super-light deck element at Krøyers Plads. Architect Cobe

CO2 production for the possibilities in a correct manner. This gives us a problem as
explained in Sect. 10.1. You cannot look up the CO2 development from something like
cement production as a single number in a table. This property varies considerably
from factory to factory, and it changes over time depending on the attention the
industry pays to developing environmentally friendly production methods.
For a traditional concrete, you produce cement in an oven with a large heat loss
fired by fossil fuel, and you do nothing to replace some of the cement with for instance
natural pozzolana. Furthermore, you apply ordinary local stone as a heavy aggregate.
For this traditional heavy concrete, you may expect to emit:

0.14 kg CO2 per kg concrete with strength 55 MPa.

The cement counts for 90% of this emission, since 1 kg concrete contains 0.14 kg
cement with 0.9 kg CO2 per kg cement. We can estimate this roughly in accordance
with [7, 11, 12].
You can then find that the cement contributes with 0.9 kg CO2 per kg cement ·
0.14 kg cement per kg concrete = 0.126 kg CO2 per kg concrete. Other contributions
(mainly transport and processing of aggregates) count for about 0.010 kg CO2 equal
to the residual 10%. In total 0.126 + 0.010 = 0.14 kg CO2 per kg concrete.
This value is also comparable with [7, 13–15].
If the concrete has a smaller compressive strength of 25 MPa, the cement content
is only about 0.10 kg per kg, but the CO2 from transport etc. is the same. You then
emit. 0.9·0.10 + 0.010 = 0.090 + 0.010 = 0.100, where [15] estimates 0.094 kg
10.2 CO2 from Materials 183

CO2 per kg concrete.

0.09 kg CO2 per kg concrete with strength 25 MPa.

The chemical process in the cement production counts for approximately 50% of
the CO2 emission [16], this is from burning limestone to calcium oxide combined
with quarts, aluminium or iron [17].

CaCO3 → CaO + CO2

You may often reduce 25% of the cement with fly ash and get:
0.75*0.9·0.14 + 0.010 = 0.095 + 0.010 = 0.105 kg CO2 pr kg 55 MPa concrete
and.
0.75*0.9·0.10 + 0.010 = 0.068 + 0.010 = 0.078 kg CO2 pr kg 25 MPa concrete.

0.11 kg CO2 per kg 55 MPa concrete with 25% fly ash or pulverized ceramic
0.08 kg CO2 per kg 25 MPa concrete with 25% fly ash or pulverized ceramic

Of this respectively 0.095 or 0.068 kg CO2 per kg concrete is from cement produc-
tion and of that 50% (equal to respectively 0.048 or 0.034 kg CO2 per kg concrete)
is from the chemical process.
If you create a new green cement replacing 35% of it with pulverized burned clay,
you may reduce the CO2 emission from the cement with 25–30%.
As mentioned in Sect. 1.1.2 this is the oldest recipe for concrete practised by the
Phoenicians more than 3000 years ago.
The Phoenicians fired their ovens with wood. In our time, we could do the same
by means of any bio fuel, so that the heating emits much less CO2 . Alternatively, we
could heat our burned limestone ovens with electricity, as we usually already do with
our ceramics ovens and glass ovens, and we could make the electricity from wind,
water, or sun power plants.
Likewise, the producer can power the mechanical processes electrically.
This means that we can reduce the CO2 emission to the one released chemically
by production of the burned limestone CaO:

0.05 kg CO2 per kg fossil free produced concrete 55 MPa

0.03 kg CO2 per kg fossil free produced concrete 25 MPa

In addition, cement hydration products gradually reabsorb atmospheric CO2

through the process of carbonation. [17] show quantities for this and how the CO2
uptake depends on the CO2 concentration, the density of the concrete, the strength
class, exposure conditions, additions, and coatings. You have not only CO2 uptake
by Calcium Hydroxide but also by the other minerals in the hydrated concrete like:

Ca(OH)2 + CO2 → CaCO3 + H2 O

184 10 Sustainability

(3CaO · 2SiO2 · 3H2 O) + 3CO2 → 3CaCO3 · 2SiO2 · 3H2 O

(2CaO · SiO2 ) + 2CO2 + xH2 O → 2CaCO3 + SiO2 · xH2 O
(3CaO · SiO2 ) + 3CO2 + xH2 O → 3CaCO3 + SiO2 · xH2 O
(3CaO · Al2 O3 · 6H2 O) + 3CO2 → 2Al(OH)3 + 3CaCO3 + 3H2 O

Pade and Guimaraes [18] show that approximately 25% of the concrete carbonize
during a 100 years lifetime, and further 25% when you demolish it.
Lo et al [19] show that application of 25% fly ash gave a marginal increase of
carbonation.
In a life cycle after carbonation, you can then estimate the CO2 release of ordinary
concrete with cement from coal fired ovens to be:

0.5 · 0.126 + 0.010 = 0.078 for ordinary 55 MPa concrete after carbonation and
0.5 · 0.090 + 0.010 = 0.055 for ordinary 25 MPa concrete after carbonation and

If all mechanical processes are made electric so that 0.010 is reduced to 0 and
25% fly ash or ceramics are applied, and the ovens are powered fossil free so that
the CO2 is only made chemically, and this is reduced by 50% carbonation you get
ultimate reduced values as

0.5 · 0.048 = 0.024 for 55 MPa and

0.5 · 0.034 = 0.017 for 25 MPa

This means that

0.024 kg CO2 per kg 55 MPa fossil free with 25% fly ash/ceramics after demolition
0.017 kg CO2 per kg 25 MPa fossil free with 25% fly ash/ceramics after demolition

This release is identical to the one made by the Phoenicians and Romans. It
should therefore be possible for us to do the same today. It is a matter of investing
in a changed production.

10.2.2 Light Aggregate Concrete

Light aggregates are usually produced by burning granulated clay in a rotary kiln.
When you heat with coal to 1150 C with a 150% heat loss from the process, you
emit

0.104 kg CO2 per MJ · 1.5 · 1150 C · (1 kJ/(kg C)) · 1 MJ/(1000 kJ) = 0.179 kg CO2

per kg light aggregate.

10.2 CO2 from Materials 185

Fig. 10.8 Light aggregate concrete blocks

A 600 kg per m3 light aggregate concrete (Fig. 10.8) has typically 400 kg expanded
clay, 100 kg sand and 100 kg cement per m3 . For a 900 kg per m3 light aggregate
concrete the numbers are 600, 150 and 150 kg per m3 . This means that the cement
and light aggregate content is proportional to the weight. The CO2 footprint per kg
is therefore the same

From cement 0.9 · 100/600 = 0.150 kg CO2 per kg concrete

+ From aggregate 0.179 · 400/600 = 0.120 kg CO2 per kg concrete
+ From sand 0.003 · 100/600 = 0.001 kg CO2 per kg concrete

In total, the emission is 0.270 kg CO2 per kg light aggregate concrete. [15] gives
0.30 kg CO2 per kg for 700 kg/m3 and 0.24 kg CO2 per kg for 1600 kg/m3 .

0.27 kg CO2 per kg concrete with expanded clay.

Typically, you replace 20% of the cement with fly ash. Doing so, you get:
From cement 0.9 · 0.8 · 100/600 = 0.120 kg CO2 per kg concrete and in total

0.120 + 0.120 = 0.240 kg CO2 per kg light aggregate concrete.

0.24 kg CO2 per kg light − aggregate concrete with 20% fly ash.

This is a typical material, and the number is the same as found by VTT for
light-aggregate concrete blocks of density 500–1600 kg/m3 [8].
The Danish light clay aggregate factory, and no doubt also other factories in the
world, is now using waste as fossil free fuel replacing 80–90% of the applied coal. The
186 10 Sustainability

last 10% of the applied carbon is embedded in the clay and cannot be replaced. With
these aggregates, you emit 0.120 from cement + 0.15*0.120 from light aggregate =
0.120 + 0.018 = 0.138 kg CO2 per kg

0.14 kg CO2 per kg concrete with light − aggregate heated by waste and 20% fly ash.

As mentioned in 10.2.1 approximately 50% of the CO2 release from cement comes
from the chemical processes and the rest from energy applied. This means that if the
cement is produced fossil free and you apply almost fossil free produced aggregate
or pumice, you only get a contribution from cement of 0.5·0.120 = 0.060 kg CO2
per kg concrete. In total

0.08 kg CO2 per kg fossil free produced light aggregate concrete with 20% fly ash.

Even light-aggregate concrete of relatively high densities of about 1800 kg/m3

carbonize significantly faster than similar heavy concrete according to for example
[19] and [20]. Carbonizing may be a problem for oxidation of steel reinforcement
[21], but for many of the applications of this book you do not use reinforcement of
steel in parts of the structure, where light aggregate concrete is applied.
Light-aggregate concrete with density less than 1200 kg/m3 is very porous and
therefore has a larger internal surface exposed to CO2 from the air and a more rapid
carbonizing time than the heavy dense concrete qualities [22].
Where we previously in this book could assess the carbonizing of a heavy dense
concrete to be 25% during its lifetime and further 25% at demolition, we can presume
it to be at least 66% in total and very often 100% in total for the light concrete qualities
of about 600 kg/m3 as we apply in super-light structures.
This means that we can assess the emission from cement to be between 0 and
0.34·0.060 = 0.020 kg CO2 per kg carbonized light aggregate concrete, and in total

0.02 − 0.04 kg CO2 per kg carbonized fossil free light

aggregate concrete with 20% fly ash

10.2.3 Pumice Concrete

If you apply volcanic material like pumice as light aggregate for a 600 kg per m3
concrete, (with 400 kg pumice, 100 kg sand and 100 kg cement per m3 replaced 20%
by fly ash), you emit from cement 0.9 · 0.8 · 100/600 = 0.120 kg CO2 per kg concrete
+ From aggregate 0 kg CO2 per kg concrete + From mechanics and transport as
for heavy concrete 0.010 kg CO2 per kg concrete = 0.130 kg CO2 per kg pumice
concrete of 600 kg/m3 and proportional the same per kg for 900 kg per m3 concrete.
10.2 CO2 from Materials 187

0.13 kg CO2 per kg 600 − 900 kg/m3 concrete with pumice and 20% fly ash

If the pumice should be transported by sea from Iceland to Denmark it costs

16 g CO2 per 1000 kg pumice per km, and you get for 1800 km an addition of
0.016·1800·400/600 =
0.019 Kg CO2 per kg pumice Concrete and in total

0.15 kg CO2 per kg 600 − 900 kg/m3 concrete with pumice and 20% fly ash in DK

10.2.4 Foam Concrete

We now look at a foam concrete made by mixing foam into a mortar of cement and
sand.
Mix design corresponds to [23]. 600 and 900 kg per m3 foam concrete with 300
and 330 kg cement per m3 . Both have replaced 20% cement by fly ash. We assess
the other contributions as similar to those for other types of concretes. The emission
is:

0.90 · 0.8 · 330/900 + 0.010 = 0.264 + 0.010 = 0.274

0.28 kg CO2 per kg foam concrete of 900 kg/m3 with 20% fly ash
0.90·0.8·300/600 + 0.010 = 0.360 + 0.010 = 0.370
0.37 kg CO2 per kg foam concrete of 600 kg/m3 with 20% fly ash

A fossil free production will then lead to

0.13 kg CO2 per kg fossil free foam concrete of 900 kg/m3 with 20% fly ash
0.18 kg CO2 per kg fossil free foam concrete of 600 kg/m3 with 20% fly ash

Also for this, you may expect more carbonation than for a heavy concrete and
you end up with between 0 and 0.34·0.18 = 0.061 kg CO2 per kg fossil free foam
concrete

0 − 0.06 kg CO2 per kg carbonized fossil free foam concrete with 20% fly ash
188 10 Sustainability

10.2.5 Aerated Concrete

Aerated concrete is produced by aluminium powder that chemically produces bubbles

with the cement. It is usually autoclaved as a part of the production process. Estimates
of the CO2 emission can be found in different sources, where different methods of
production presumably cause the different values [8, 15]:

0.51 kg CO2 per kg aerated reinforced concrete block 433 kg/m3

0.34 kg CO2 per kg aerated reinforced concrete 480 kg/m3

10.2.6 Steel

Steel production consists of several processes each giving rise to release of CO2
partly from heating and partly from the chemical reactions taking place in the material
(Fig. 10.9).
At first, you make coke from coal by heating it above 600 ºC in absence of oxygen.
Then, you make pig iron (also called crude iron) by melting at 1600 ºC the raw
materials of iron ore Fe2 O3, coke C and lime CaCO3 and impurities such as ferrous
sulphide FeS in a blast furnace.

2Fe2 O3 + 6C + 3O2 → 4Fe + 6CO + 3O2 → 4Fe + 6CO2 and

2FeS + 2CaCO3 + C → 2CaS + 2FeO + 2CO2 + C → 2CaS + 2Fe + 3CO2

Fig. 10.9 Firth of forth bridge by Fowler and Baker 1889. Photo KD Hertz
10.2 CO2 from Materials 189

The pig iron has a carbon content of about 4% and silica and other impurities
making it too brittle for application in structures.
Then, the primary steel process reduces the content of carbon to about 1%.

C + O2 → CO2

For that, you apply a Basic Oxygen Furnace or an Electric Arc Furnace, which
has replaced the old Bessemer furnace, because they have better pollution control
systems.
In this process, you also want to remove a surplus of oxygen by adding aluminium.

4Al + 3O2 → 2Al2 O3 .

Then in the so-called secondary steel process, you adjust the composition of the
molten steel.
All these processes of traditional steel production lead to a relatively large CO2
emission mainly caused by chemical reactions. You can assess the release as

3.00 kg CO2 per kg virgin steel for the traditional production methods
2.80 kg CO2 per kg virgin steel and wrought iron for modern production methods

This is according to for example [7, 15]. Germeshuizen and Blom [24] and others
indicate a possibility of reducing the CO2 emission from steel productions drastically
by application of hydrogen replacing carbon as reducing agent.

3Fe2 O3 + 9H2 → 2Fe3 O4 + H2 O + 8H2 → 6FeO + 3H2 O + 6H2 → 6Fe + 9H2 O

They estimate that this may reduce the CO2 emission to

0.18 kgCO2 per kg virgin steel for a future hydrogen based steel production

The steel industry still has to develop and implement this. One of the problems
for this is to avoid excessive hydrogen in the steel that makes it brittle. Therefore,
you may consider this only to be a possible future value.

10.2.7 Recycled Steel

Hammond and Jones [7] also estimate a value for recycled steel as 0.47 kg CO2 per
kg, where [15] estimates 0.73 kg CO2 per kg probably due to different production
methods. These differences and differences in the fraction of reused steel that is
applied on average leads to different values for the steel usually applied in Germany
and UK.
190 10 Sustainability

With the average content of recycled steel in UK, [7] estimate

1.46 kg CO2 per kg steel with average recycled content in UK

With the average content of recycled steel in Germany, [15] estimates

1.00 kg CO2 per kg steel profile with average recycled content in Germany

Applying the same share of recycling as used today, you get for a future recycled
hydrogen based steel 0.18·1.46/2.80 =

0.094 kg CO2 per kg steel for future recycled hydrogen based steel,

If a steel profile is galvanized, this protection increases the CO2 impact.

According to [15] you emit

1.85 kg CO2 per kg galvanized steel profile

Stainless steel is a collective name for a number of alloys with different materials.
Their CO2 impacts are therefore varying according to the content and the methods
of production. You may find values like [7, 8]

3.78 kg CO2 per kg stainless steel according to Ruuska [8]

6.15 kg CO2 per kg stainless steel according to Hammond and Jones [7]

Reinforcing steel
Reinforcing steel with an average recycled content gives according to [15]

0.75 kg CO2 per kg reinforcement

Typically, you apply 100 kg reinforcement per m3 concrete increasing the density
from 2300 to 2400 kg per m3 . This means that a typical value for reinforced concrete
can be found by adding the CO2 impact from the reinforcement to that of the concrete
and relate the sum to the new larger density.
Using an average value for concrete of 0.13 kg CO2 per kg, this will give you an
addition of

100 kg·0.75 kg CO2 per kg/2400 kg−0.13 ·(2400 − 2300)/2400 = 0.031−0.005

= 0.03 kg CO2 addition per kg concrete from 100 kg reinforcement per m3
10.2 CO2 from Materials 191

10.2.8 Timber

Some politicians believe that wood and timber structures (Fig. 10.10) represent a
solution to the climate crisis, because trees absorb CO2 when they grow. However,
the trees release the CO2 again, when they stop growing and start to rot, burn, or are
eaten by insects.
When you apply wood in a structure, you have to prevent it from rotting or
burning by taking special precautions and often by impregnation or painting with
poisonous chemicals at regular intervals. The more CO2 you store in wood, the more
comprehensive is the task that you hand over to the next generations, preventing the
wood from rotting or burning and releasing the old CO2 .
The authors consider this behaviour to be the opposite of saving the next gener-
ation from our CO2 problems. Instead, we claim that it is a better solution to the
climate problems, if you reduce the release of greenhouse gases to a minimum and
build structures with a long lifetime in order to save CO2 from material production,
transport, and building processes.
At present, timber structures give rise to CO2 footprints from the processes of
forestry, sawmill work, transport, construction, maintenance, and demolition.
In the following, we try to find representative values by comparing different
sources estimating the CO2 emission.
Most data sources give values from phase A, B and C representing the full life
cycle (Figs. 10.5 and 10.6). However, the much-used reference Ökobaudat [15] gives
a separate value for CO2 emission from phase A comprising the materials, making
the components, and constructing of the building. For wood structures, this becomes
a negative number, since trees absorb CO2 when they grow. For example, Ökobaudat

Fig. 10.10 Copenhagen’s first Central Station by Herholdt 1863. Photo KD Hertz
192 10 Sustainability

gives a value of −632 kg CO2 per m3 CLT (Cross Laminated Timber) that with a
density of 489 kg/m3 gives −1.291 kg CO2 /kg CLT.
However, you release the accumulated CO2 again when the material burns or rots.
Ökobaudat [15] includes a value for this release in phase C “End of Life” as
1.802 kg CO2 per kg wood. We need to add this value to the negative value of phase
A and B in order give the impact on the climate and to compare it to values from
other sources representing phase A, B, and C. This gives a total of 0.51 kg CO2 per
kg CLT, which is comparable to values from other data sources.
Sometimes, you hear an argument, that new trees are planted and they absorb the
CO2 released by the old. However, this argument makes no sense, since the new trees
also releases their CO2 .
In order to investigate this further we make a separate check of the value for
phase C. We compare the value applied above for phase C with the same found by
other references. Using a standard value for CO2 release from burning of wood (for
example from Quasching [25] or Engineering Toolbox [26], you get 0.108 kg CO2
per MJ.
References like Aniszewska and Arkadiusz [27] and Alakangas [28] find that at
12% moisture, the energy released by burning is about 16 MJ/kg increasing for a
smaller moisture content, where the density also decreases.
This gives a value of 16 ·(MJ/kg) · 0.108 kg CO2 /MJ = 1.73 kg CO2 /kg, which
indicates that the value 1.80 kg CO2 per kg wood for phase C from [15] comprising
demolition and incineration in a municipal solid waste plant appears to be reasonable.
We have therefore added this value for phase C to all values for phase A1-3 from
Ökobaudat. As explained, this gives 0.51 kg CO2 per kg CLT as shown below.

0.51 kg CO2 per kg cross laminated timber (CLT)

Lifetime of timber structures is in general smaller than for stone structures due
to rot, fire, and insects. However, reuse of unharmed timber is more common, and
these two opposite effects makes it reasonable to compare CO2 emissions for equal
life times.
For ordinary timber, different sources may give slightly different values as for
example.
0.31 Kg CO2 per kg timber in [7] and 0.38 kg CO2 per kg sawn pine timber of
density 549 kg/m3 in [15].
We apply average values

0.35 kg CO2 per kg timber

0.52 kg CO2 per kg laminated wood, plywood, or chipboard
10.2 CO2 from Materials 193

10.2.9 Other Materials

Usually, you apply a number of other materials in structures or as a part of them

protecting them from impact of fire or weather.
For some of them as for example aluminium, you have the same issue of recycling,
as we saw for steel that virgin and recycled aluminium have quite different CO2
emissions, and you may take an average degree of recycling into account [7].

12.50 kg CO2 per kg virgin aluminium profile

2.12 kg CO2 per kg recycled aluminium profile
10.93 kg CO2 per kg aluminium profile with average recycling

A data sheet at the end of the book summarises the CO2 data given in this chapter
and provides data for a number of other materials as gypsum board, brick, glass,
aluminium, mineral wool and foam plastic.

10.3 CO2 from Processes and Transport

10.3.1 CO2 from Heating

Heating with oil, diesel, or gas gives rise to

0.24 kg CO2 per kWh = 0.24/3.6 = 0.067 kg CO2 per MJ energy.

Heating with coal or wood gives rise to

0.39 kg CO2 per kWh = 0.37/3.6 = 0.108 kg CO2 per MJ energy.

10.3.2 CO2 from Transport

As long as we base transport on fossil fuel, you need to estimate the transport needed
for raw materials, building materials and building elements.
The following average values are based on the work made by [29], and reported
by [30]

8 g CO2 for 1000 kg per km ship deep − sea container

16 g CO2 for 1000 kg per km ship short sea
31 g CO2 for 1000 kg per km barge
194 10 Sustainability

22 g CO2 for 1000 kg per km rail

(Average, 35 diesel hauled, 18 electric hauled. ECTA)
62 g CO2 for 1000 kg per km road
602 g CO2 for 1000 kg per km air

10.4 CO2 from Building Components

10.4.1 Building Components with 55 dB Sound Insulation

A main hindrance for reducing CO2 emission producing domestic building compo-
nents is the consideration of sound insulation. A good sound insulation reducing
noise at low frequencies requires that the separating building component contain a
certains mass.
In the following, we analyze the total CO2 emission for decks and walls fulfilling
the Danish requirement of 55 dB sound insulation.
To obtain that with a massive structure of one material, you have to apply 440 kg
per m2 wall or deck structure.
In super-light deck elements (SL-decks) as the one in Fig. 10.11, you combine
layers of materials of quite different stiffness and eigenfrequency such as heavy
strong concrete of density 2300 kg per m3 and a light concrete of density 700 kg per
m3 . The two materials oscillate differently, and some of the sound reduces to heat.

Fig. 10.11 SL-deck element for DTU Building 130. Photo KD Hertz
10.4 CO2 from Building Components 195

This is the reason why you may reduce the total mass required for these components
to about 340 kg per m2 .

10.4.2 Building Decks with 55 dB Sound Insulation

190 mm massive concrete deck

A 190 mm slab of 55 MPa concrete with for example five pretensioned 12.5 mm
wires per m emits 437 kg concrete of 0.14 kg CO2 per kg and 2.7 kg reinforcement
per m2 . In total:

63.8 kg CO2 per m2 190 mm massive concrete deck

220 mm hollow-core slab with additional 55 mmtop concrete

A 220 mm hollow core deck of 55 MPa concrete has in an actual design 6 holes
of diameter 147 mm per 1.2 m width. This gives a concrete mass of 311 kg per m2
and to that 3.65 kg per m2 from the prestressing reinforcement. Therefore, it requires
a 55 mm top concrete of 126 kg per m2 , which can be made from a 25 MPa quality.
This gives 44.0 + 11.3 + 2.7 =

58.1 kg CO2 per m2 220 mm hollow − core slab with 55 mm top concrete

180 mm super-light SL-deck with 20 mm top concrete.

It uses the five pretensioned wires per m, 233 kg per m2 strong concrete of 55 MPa
and 58 kg per m2 light aggregate concrete of 700 kg/m3 .
It can obtain the 55 dB sound insulation with a total weight of 340 kg per m2 .
20% of the cement is replaced with fly ash and you get 0.24 kg CO2 per kg light
concrete.
In Denmark and some other places, light aggregates are made in an 85% fossil
free (FF) process. For these materials, you emit 0.14 kg CO2 per kg light concrete.
CO2 emission from light concrete per m2 then becomes 58 kg · 0.24 = 13.9 kg
CO2 (FF 8.1 kg CO2 ). To that, you should add 233 kg strong concrete of 55 MPa
and 0.14 kg CO2 /kg (giving 32.6 kg CO2 per m2 ) + 45 kg top concrete of 25 MPa
and 0.09 kg CO2 /kg (giving 4.1 kg CO2 per m2 ) + 2.7 kg CO2 per m2 for the
reinforcement.
In total, you emit 32.6 + 13.9 + 4.1 + 2.7 =

53.3 kg CO2 per m2 180 mm SL deck with 20 mm top concrete

With 85% fossil free light aggregates, you emit 32.6 + 8.1 + 4.1 + 2.7 =

47.5 kg CO2 per m2 180 mm SL deck with 85% fossil free

light aggregates and with 20 mm top concrete
196 10 Sustainability

175 mm CLT five-layer slab with 110 mm top concrete

A 175 mm five-layer CLT slab of density 489.4 kg/m3 has a mass of 85.6 kg per
2
m.
You should provide it with a 110 mm top concrete of 254 kg/m2 in order to reach
the 340 kg/m2 that you at least should apply to reach the required 55 dB sound
insulation for a composite structure.
It emits 85.5 · 0.51 = 43.7 from the CLT + 254 · 0.09 = 22.9 from the concrete
=

66.6 kg CO2 per m2 five − layer CLT deck with 110 mm top concrete

10.4.3 Sand as Sound Insulator for Decks

340 kg per m2 required for composite sound insulation has been obtained by
adding sand for example to a CLT deck. However, it gave practical problems with
sand moving and even penetrating the structure, when the deck was subjected to
oscillations.
Therefore, the contractors known to have applied it, now use a concrete topping
instead.
Nevertheless, if you calculate this solution, you emit for a five layer CLT slab of
86 kg per m2

43.7 kg CO2 per m2 five − layer CLT with additional 254 kg

sand per m2 (not a pragmatic solution)

If you compare this with a 90 mm pretensioned massive concrete slab of 55 MPa

concrete of 210 kg per m2 with additional 130 kg sand per m2 , the impact becomes

31.2 kg CO2 per m2 .90 mm massive concrete slab

+ additional 130 kg sand per m2 (not a pragmatic solution)

10.4.4 Building Walls with 55 dB Sound Insulation

190 mm massive concrete wall

437 kg 25 MPa concrete of 0.09 kg CO2 per kg emits 39.3 kg CO2 per m2 wall.
Two layers of 6 mm reinforcement per 250 mm in each direction give 3.55 kg
reinforcement of 0.75 kg CO2 per kg and this emits 2.7 kg CO2 per m2 wall.
The wall has a total weight of 440 kg/m2 , and the emission is
10.4 CO2 from Building Components 197

42.0 kg CO2 per m2 massive 190 mm concrete wall

Massive CLT wall

For a massive timber wall the emission would be similar to a massive timber slab.
For a massive wall or slab to obtain 55 dB sound insulation, you should
apply 440 kg/m2 , and if that is CLT (Cross Laminated Timber) with a density of
489.4 kg/m3 , this requires a thickness of 899 mm emitting

224.4 kg CO2 per m2 massive CLT wall

You can also, for walls with 55 dB sound insulation, benefit from making
composite structures such as the examples shown below.
180 mm sandwich concrete wall
With two layers of 65 mm normal concrete of 25 MPa of 299 kg/m2 and 50 mm
light concrete of 35 kg per m2 and two reinforcing nets. The total weight is 340 kg
per m2 and that gives 55 dB since 360 kg has proven by test to give 58 dB with the
same relative proportions between light and heavy concrete.
Using light aggregate concrete of 0.24 kg CO2 per kg you emit 26.9 from the
normal concrete + 8.4 from the light + 2.7 from the reinforcement =

38.0 kg CO2 per m2 sandwich wall of 50 mm

light and 2 · 65 mm normal concrete

With 85% fossil free light aggregates with 0.14 kg CO2 per kg light concrete, the
result is

26.9 + 4.9 + 4.1 + 2.7 =

34.5 kg CO2 per m2 sandwich concrete wall with 85%
fossil free light aggregates

Composite CLT walls were tested and calculated by [31].

The best structure fulfilling a sound insulation requirement of 55 dB was:
CLT wall of two slabs with a void between
The slabs are each 120 mm, and the void of 95 mm can be filled with mineral
wool.
It has a minimal total weight of only 171 kg/m2 and emits

87.4 kg CO2 per m2 CLT sandwich wall of 2 × 120 mm

+ 95 mm mineral wool
198 10 Sustainability

10.4.5 Building Components with Other Requirements

100 mm massive prefabricated concrete deck (not for domestic buildings)

230 kg concrete of 55 MPa (0.14 kg CO2 per kg) emits 32.2 kg CO2 per m2 deck.
In addition 3.65 kg reinforcement emits (0.75 kg CO2 per kg) 5.3 kg CO2 per m2
deck.
(3.65 kg is equal to five 12.5 mm prestressing lines per m, each consisting of
93 mm2 of steel with density 7850 kg/m3 ).
The total weight of such deck is 234 kg/m2 . Note that this deck is not sufficient
for domestic buildings in general because a sound insulation requirement of 55 dB
would require a mass of 440 kg/m2 . However, it is applied for industrial buildings.

34.9 kg CO2 per m2 100 mm massive prefabricated

concrete deck for industry

150 mm massive concrete wall

345 kg concrete of 25 MPa (0.09 kg CO2 per kg) emits 31.1 kg CO2 per m2
wall- Two layers of 6 mm reinforcement per 250 mm in each direction gives 3.55 kg
reinforcement (0.75 kg CO2 per kg) and this emits 2.7 kg CO2 per m2 wall. The total
weight of the wall is 349 kg/m2 .

33.7 kg CO2 per m2 massive 150 mm concrete wall for industry

200 mm massive concrete wall (with 58 dB sound insulation)

460 kg of 25 MPa concrete (0.09 kg CO2 per kg) emits 41.4 kg CO2 per m2
wall. Two layers of 6 mm reinforcement per 250 mm in each direction gives 3.55 kg
reinforcement per m2 wall (0.75 kg CO2 per kg), and this emits 2.7 kg CO2 per m2
wall.
The wall has a total weight of 464 kg/m2 , and gives an emission of

44.1 kg CO2 per m2 massive 200 mm concrete

wall with 58 dB sound insulation

220 mm SL-deck (with 58 dB sound insulation)

The 71 kg light aggregate concrete (0.24 kg CO2 per kg) emits 17.0 kg CO2 per
m2 deck. The 285 kg normal concrete of 55 MPa (0.14 kg CO2 per kg) emits 39.9 kg
CO2 per m2 deck. The 3.65 kg reinforcement (0.75 kg CO2 per kg) emits 2.7 kg
CO2 per m2 deck. (3.65 kg is equal to five 12.5 mm prestressing lines per m each
consisting of 93 mm2 with steel density 7850 kg/m3 ). The slab proved to give about
58 dB sound insulation, which is more than the required 55 dB. The total weight of
the deck is 360 kg/m2 , and it emits 39.9 + 17.0 + 2.7 =

59.7 kg CO2 /m2 220 mm SL − deck with 58 dB sound insulation

10.4 CO2 from Building Components 199

With 85% fossil free light aggregates the emission is 39.9 + 9.9 + 2.7 =

52.6 kg CO2 per m2 220 mm SL − deck with 85%

fossil free light aggregates and 58 dB sound insulation

180 mm SL-deck
It has 58 kg light concrete that emits 13.9 kg CO2 per m2 deck and 233 kg normal
55 MPa concrete that contributes with 32.6 kg CO2 per m2 deck. Furthermore, an
emission of 2.7 kg CO2 per m2 deck comes from 3.7 kg reinforcement.
The total weight of the deck is 295 kg/m2 , and it emits 32.6 + 13.9 + 2.7 =

49.3 kg CO2 /m2 180 mm SL − deck for industry

With 85% fossil free light aggregates, it becomes 32.6 + 8.1 + 2.7 =

43.5 kg CO2 per m2 180 mm SL − deck with 85%

fossil free light aggregates

200 mm concrete sandwich wall (with 58 dB sound insulation)

The sandwich wall consists of two layers of 72 mm normal concrete of 2300 kg/m3 ,
and one layer of 56 mm light concrete of 700 kg/m3 . The light concrete contributes
with 39 kg and 9.3 kg CO2 per m2 (0.24 kg CO2 per kg). In addition, the normal
concrete of 55 MPa weights 332 kg and emits (0.09 kg CO2 per kg) 29.9 kg CO2 per
m2 wall. 3.6 kg reinforcement emits 2.7 kg CO2 per m2 deck. (3.6 kg is equal to two
nets of 6 mm bars per 250 mm with density 7850 kg/m3 ). This gives a total weight
of 375 kg/m2 and approximately 60 dB sound insulation. The emission is 29.9 + 9.3
+ 2.7 =

41.9 kg CO2 per m2 200 mm sandwich wall of 50 mm

light and 2 · 65 mm normal concrete and 58 dB sound insulation

With 85% fossil free light aggregates, the emission is.

(0.14 kg CO2 per kg) 29.9 + 5.4 + 2.7 =

38.0 kg CO2 per m2 200 mm sandwich wall of 50 mm light 85%

fossil free concrete and 2 · 65 mm normal
concrete and 58 dB sound insulation

150 mm sandwich concrete wall (with 45 dB sound insulation)

The sandwich wall consists of two layers of 54 mm normal concrete of 2300 kg/m3
and one layer of 42 mm light concrete of 700 kg/m3 . The light concrete contributes
with 29 kg weight and 7.0 kg CO2 per m2 (0.24 kg CO2 per kg). In addition, the
normal concrete of 55 MPa weights 249 kg and emits (0.09 kg CO2 per kg) 22.4 kg
200 10 Sustainability

CO2 per m2 wall. 3.6 kg reinforcement emits 2.7 kg CO2 per m2 deck. (3.6 kg is
equal two nets of 6 mm bars per 250 mm with density 7850 kg/m3 ). This gives a
total weight of 282 kg/m2 and approximately 45 dB sound insulation. The emission
is 22.4 + 7.0 + 2.7 =

32.1 kg CO2 per m2 150 mm sandwich wall of 42 mm

light concrete and 2 ∗ 54 mm normal concrete and 45 dB
sound insulation for industry

With 85% fossil free light aggregates with 0.14 kg CO2 per kg 22.4 + 4.1 + 2.7
=

29.2 kg CO2 per m2 150 mm sandwich wall of 42 mm

light 85% fossil free concrete and 2 ∗ 54 mm
normal concrete and 45 dB sound insulation for industry

10.4.6 Sustainability

The numbers in this chapter demonstrate how you can save CO2 emission by appli-
cation of the new technology presented in the book, when you compare components
with the same functional requirements. They also show possibilities for obtaining
further reductions in the future by application of improved production methods
already known today.
The applied materials such as lime, clay, sand, and water are clean and non-toxic,
and methods for eliminating the application of carbon for production are already
implemented in some factories.
The authors therefore hope, that the technology presented in the book can
contribute to a more climate-friendly and sustainable building industry, which has
also been evaluated by independent price committees.

References

1. Andersen S (1979) The accumulated energy consume for building materials. Report 134, (in
Danish) DTU Byg (Inst.of Building Design) 1979, 198p
2. Andersen S (1980) The accumulated energy consume for dwellings. Report 137, (in Danish)
DTU Byg (Inst.of Building Design) 1980, 124p
3. Byrne C (2010) VentureBeat San Francisco The New York Times November 17, 2010
4. CEN (2012) EN 15804 Sustainability of construction works. Environmental product declara-
tions, +A2(2019):66p
References 201

5. Herrmann H (2017) Ökobaudat Basis for the building life cycle assessment. Forshung für
die Praxis Volume 11. Federal Insitute for Research on Building, Urban Affairs and Spatial
Development. Bonn, 40p
6. Kupfer T et al (2019) GaBi database and modelling principles
7. Hammond G, Jones C (2011) Inventory of carbon and energy (ICE). University of Bath Version
2.0. http://www.circularecology.com/embodied-energy-and-carbon-footprint-database.html#.
XbbGPuhKhaR
8. Ruuska A (2013) Carbon footprint for building products. ECO2 data for materials and products
with the focus on wooden building products. VTT Technology 115. VTT
9. Benhelal E et al (2013) Global strategies and potentials to curb CO2 emissions in cement
industry. J Clean Prod 51:142–161
10. Gregg JS et al. (2008) China: Emissions pattern of the world leader in CO2 emissions from
fossil fuel consumption and cement production. Geophys Res Lett Adv Earth Space Sci 35(8)
11. NRMCA (2008) Concrete CO2 Fact Sheet. Publication Number 2PCO2. National Ready Mixed
Concrete Association
12. Pentalla V (1997) Concrete and sustainable development. ACI Mater J 94(5):409–416
13. Hertz KD (2010) CO2 emissions from super-light structures. Department of Civil Engineering,
Technical University of Denmark, 6p, 2010
14. Hertz KD, Bagger A (2011) CO2 emissions from super-light structures. Proceedings of the
IABSE-IASS Symposium “Taller, Longer, Lighter” 7p, London 20–23 Sep 2011
15. Ökobaudat (2020) Database. German Federal Ministry of the Interior, Building and Community
(BMI). https://www.oekobaudat.de/datenbank/browser-oekobaudat.html
16. Lehne J, Preston F (2018) Making concrete change—Innovation in low-carbon cement and
concrete. Chatham House, Royal Institute of International Affairs, 122p. https://www.cha
thamhouse.org/publication/making-concrete-change-innovation-low-carbon-cement-and-con
crete
17. Xi F et al (2016) Substantial global carbon uptake by cement carbonation. Nature Geoscience,
vol 9. Macmillan Publishers Ltd
18. Pade C, Guimaraes M (2007) The CO2 uptake of concrete in a 100 year perspective. Cement
Concr Res 37:1348–1356
19. Lo TY et al (2007) Comparison of carbonation of lightweight concrete with normal weight
concrete at similar strength levels. Constr Build Mater 22:1648–1655
20. Yingli G et al (2013) Effects of different mineral admixtures on carbonation resistance of
lightweight aggregate concrete. Constr Build Mater 43:506–510
21. Lagerblad B (2005) Carbon dioxide uptake during concrete life cycle—State of the art. Report
2:2005 Swedish Cement and Concrete Research Insitute, 47p
22. Goltermann P et al (2017) Climate friendly concrete. (In Danish: Klima venlig Beton) For
Region H. Department of Civil Engingeering, Technical University of Denmark. BYG R-371,
53p
23. SirContec (2010) www.sircontec.com
24. Germeshuizen LM, Blom PWE (2013) A techno-economic evaluation of the use of hydrogen
in a steel production process, utilizing nuclear process heat. Int J Hydrogen Energy 38:10671–
10682
25. Quasching V (2019) Specific carbon dioxide emissions of various fuels. Erneubare energien
und klimaschutz, Berlin. https://www.volker-quaschning.de/
26. Engineering Toolbox (2019). https://www.engineeringtoolbox.com/
27. Aniszewska M, Arkadiusz G (2014) Comparison of heat of combustion and calorific value of
the cones and wood of selected forest trees species. Lesne Prace Badawcze (Forest Research
Papers) 75(3):231–236
28. Alakangas E (2005) Properties of fuels used in Finland—BIOSOUTH-project. VTT processes
report PRO2/P2030/05 technical research centre of Finland, 100p
29. McKinnon A, Piecyk M (2011) Measuring and managing CO2 emissions. Logistics Research
Centre, Hariot-Watt University, Edinburgh, UK, 36p. https://cefic.org/app/uploads/2018/12/
MeasuringAndManagingCO2EmissionOfEuropeanTransport-McKinnon-24.01.2011-REP
ORT_TRANSPORT_AND_LOGISTICS.pdf
202 10 Sustainability

30. ECTA (2011) Guidelines for measuring and managing CO2 emission from freight transport
operations. European Chemical Transport Association. Brussels, 18p. https://www.ecta.com/
resources/Documents/Best%20Practices%20Guidelines/guideline_for_measuring_and_man
aging_co2.pdf
31. Ljunggren F (2019) Sound insulation prediction of single and double CLT panels. In: (Luleå
Univ.) Proceedings of 23rd international congress on acoustics, Aachen, pp 242–248
Chapter 11
CO2 Data

Materials
CO2 emissions per 2020 from materials including a full life cycle with production,
transport, building, application, demolition and removal.

Steel virgin from factory and wrought iron 2.80 kg CO2 /kg
Steel recycled 0.47 kg CO2 /kg
Steel profiles with average recycled content in Germany 1.00 kg CO2 /kg
Steel profile hot galvanized 1.85 kg CO2 /kg
Reinforcement with average recycled content 0.75 kg CO2 /kg
100 kg Reinforcement in 2400 kg/m3 concrete gives an addition of 0.03 kg CO2 /kg
Cement 0.90 kg CO2 /kg
Concrete 55 MPa 0.14 kg CO2 /kg
Concrete 25 MPa 0.09 kg CO2 /kg
Light concrete (600–900 kg/m3 ) 0.27 kg CO2 /kg
Light concrete with 20% fly ash 0.24 kg CO2 /kg
Light concrete with 20% fly ash and 85% fossil free aggregates 0.14 kg CO2 /kg
Aerated concrete 480 kg/m3 on aluminium powder 0.34 kg CO2 /kg
Timber 0.35 kg CO2 /kg
CLT Cross-Laminated Timber 489 kg/m3 0.51 kg CO2 /kg
Laminated wood, plywood, chipboard 507 kg/m3 0.52 kg CO2 /kg
Gypsum plaster board 630 kg/m3 0.23 kg CO2 /kg
Brick 1800 kg/m3 0.23 kg CO2 /kg
Glass 3 mm 2500 kg/m3 1.37 kg CO2 /kg
Aluminium profile 10.93 kg CO2 /kg
Mineral wool 1.60 kg CO2 /kg
Foam Plastic PE 6.13 kg CO2 /kg

Orign of data
Anybody can look the presented data up in references like ÖkobauDat [7] or LCA
Byg [8] and you would get almost the same data from other references like Hammond
and Jones [3] or Ruuska VTT [5].
© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 203
K. D. Hertz and P. Halding., Sustainable Light Concrete Structures, Springer Tracts
in Civil Engineering, https://doi.org/10.1007/978-3-030-80500-5_11
204 11 CO2 Data

The data represent the full life cycle including the end of life stage, which is
important for the impact on the climate. End of life is especially important for organic
materials like wood that absorbs CO2 when it grows, and releases it at the end of life.
Sources like [3] and [5] do therefore not present the absorbed CO2 without the end
of life contribution in their numbers. However, Using ÖkobauDat [7] or LCA Byg
[8], you have to add end of life as explained in detail in [1] and [6]. For example, [7]
gives a value of −632 kg CO2 /m3 for Cross Laminated Timber CLT. With density
489 kg/m3 it is −1.291 kg CO2 /kg. To that, you should add a value of 1.80 kg CO2 /kg
for end of life, so that you get 0.51 kg CO2 /kg for CLT. This is close to the values
from the other sources.

Transport

8 g CO2 for 1000 kg per km ship deep sea container

16 g CO2 for 1000 kg per km ship short sea
31 g CO2 for 1000 kg per km barge
22 g CO2 for 1000 kg per km rail
(1 km electric per 1.9 km diesel hauled. ECTA)
62 g CO2 for 1000 kg per km road
602 g CO2 for 1000 kg per km air

Energy
Diesel oil or Gas 0.24 kg CO2 /kWh = 0.067 kg CO2 /MJ.
Coal or Wood 16 ·(MJ/kg) · 0.108 kg CO2 /MJ = 1.73 kg CO2 /kg.

Building structures with 55 dB sound insulation

190 mm Massive concrete deck 55 MPa 440 kg/m2 63.8 kg CO2 /m2
220 mm Hollow-core + 126 kg 55 mm top concrete 440 kg/m2 58.1 kg CO2 /m2
180 mm SL-deck + 45 kg 20 mm top concrete 53.3 kg CO2 /m2
180 mm SL-deck with 85% fossil free aggr + 45 kg top concrete 47.5 kg CO2 /m2
220 mm SL-deck 360 kg/m2 . (makes 58 dB) 59.7 kg CO2 /m2
220 mm SL-deck with 85% fossil free aggr (makes 58 dB) 52.6 kg CO2 /m2
900 mm Massive CLT deck or wall 440 kg/m2 224.4 kg CO2 /m2
175 mm CLT deck 5-ply 86 kg/m2 with 110 mm top concrete 66.6 kg CO2 /m2
Steel deck 150 kg/m2 with 290 kg top concrete per m2 303.6 kg CO2 /m2
190 mm massive concrete wall 440 kg/m2 42.0 kg CO2 /m2
180 mm sandwich concrete wall 340 kg/m2 38.0 kg CO2 /m2
180 mm sandwich concrete wall with 85% fossil free aggr 34.5 kg CO2 /m2
CLT wall as tested 2·120 mm CLT + 95 mm air 171 kg/m2 87.4 kg CO2 /m2
References 205

References

1. CEN (2012) EN 15804 Sustainability of construction works. Environmental product declara-

tions. + A2 (2019) 66p
2. ECTA (2011) Guidelines for measuring and managing CO2 emission from freight transport
operations. European Chemical Transport Association. Brussels, 18p
3. Hammond G, Jones C (2011) Inventory of carbon & energy (ICE). University of Bath
Version 2.0. http://www.circularecology.com/embodied-energy-and-carbon-footprint-database.
html#.XbbGPuhKhaR
4. Hertz KD, Halding P (2021) Sustainable light concrete structures
5. Ruuska A (2013) Carbon footprint for building products. ECO2 data for materials and products
with the focus on wooden building products. VTT Technology 115. VTT
6. Hermann H (2017) ÖkobauDat – Zukunft Bauen. Forschung für die Praxis Vol 11. 40p
7. Ökobaudat (2019) Free Database. German Federal Ministry of the Interior, Building and
Community (BMI). OBS! Remember to add phase C demolition to A + B for full cycle. https://
www.oekobaudat.de/datenbank/browser-oekobaudat.html
8. LCA Byg 4.0 Beta (2020) Free LCA program based on Ökobaudat version 2016. https://www.
lcabyg.dk/
9. Ljunggren F (2019) Sound insulation prediction of single and double CLT panels. Proceerings
of the 23rd Intermational Concress on Acoustics, Sept. 2019 Aachen, Germany, pp 242–248