0% found this document useful (0 votes)

124 views

Aci sp-224-2004

Uploaded by

Abigael Valles Ramirez

Available Formats

Download as PDF, TXT or read online on Scribd

0% found this document useful (0 votes)

124 views

Aci sp-224-2004

Uploaded by

Abigael Valles Ramirez

Available Formats

Download as PDF, TXT or read online on Scribd

You are on page 1/ 220

daneshlink.

com

THIN REINFORCED CEMENT-BASED

PRODUCTS AND
CONSTRUCTION SYSTEMS

American Concrete Institute®

Editor Advancing concrete knowledge
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute t " n nnA

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced
Cement-Based Products and
Construction Systems

Editor
Ashish Dubey
American Concrete lnstitutelt
Advancing concrete knowledge

SP-224
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

First printing, November 2004

DISCUSSION of individual papers in this symposium may be submitted in accordance

with general requirements of the ACI Publication Policy to ACI headquarters at the address
given below. Closing date for submission of discussion is May 2005. All discussion approved
by the Technical Activities Committee along with closing remarks by the authors will be
published in the September/October 2005 issue of either ACI Structural Journal or ACI
Materials Journal depending on the subject emphasis of the individual paper.

The Institute is not responsible for the statements or opinions expressed in its publications.
Institute publications are not able to, nor intended to, supplant individual training,
responsibility, or judgment of the user, or the supplier, of the information presented.

The papers in this volume have been reviewed under Institute publication procedures by
individuals expert in the subject areas of the papers.

All rights reserved, including rights of reproduction and use in any form or by any means,
including the making of copies by any photo process, or by any electronic or mechanical
device, printed or written or oral, or recording for sound or visual reproduction or for use in
any knowledge or retrieval system or device, unless permission in writing is obtained from
the copyright proprietors.

Cover Page Photograph

The Clubhouse at Abu Dhabi - The building demonstrates the use of thin cementitious
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

panels as exterior cladding.

Photograph courtesy of Fibrex LLC, Abu Dhabi

Printed in the United States of America

Editorial production: Lindsay K. Kennedy

Library of Congress catalog card number: 2004114457

ISBN: 0-87031-159-X

Copyright American Concrete Institute

PREFACE

The use of thin reinforced cementitious products continues to grow rapidly today in a
variety of construction applications worldwide. Thin reinforced cementitious products are
strong and tough, dimensionally stable, fire resistant, and possess excellent moisture
resistance and environmental durability. With the continued rapid decline in the use of
asbestos fibers as reinforcement in thin cementitious products, several advancements have
occurred in the field as a result of the development of new types and forms of reinforcement
for thin cementitious products. Also, several advancements have taken place in the material
science and manufacturing methods of cementitious materials for thin reinforced
cementitious products. This publication contains the papers originally presented in a
symposium on the topic of thin reinforced cementitious products organized by ACI
Committee 549 on Thin Reinforced Cementitious Products and Ferrocement during the ACI
2003 Spring Convention held in Vancouver, Canada. The symposium explored current
state-of-the-art and recent advances in material science, manufacturing methods, and
practical applications of thin reinforced cementitious products.

The topics covered in this publication include material science oftextile reinforced concrete,
use of textile reinforced concrete for integrated formwork and exterior cladding panels,
prestressed thin-sheet concrete products, ultra-high-performance thin precast concrete
products, production of concrete tubes by centrifugation method, freezing-and-thawing
durability of commercial fiber-reinforced cement boards, structural evaluation of cement-
skin sandwich building systems, microwave accelerated curing method for producing precast
cementitious products, history of glass fiber-reinforced concrete (GFRC) products, and
modeling of cement-based laminate composites. The papers presented in this publication
have been peer reviewed by experts in the field, according to the guidelines established by
the American Concrete Institute.

The future of thin reinforced cementitious products depends largely on their ability to
compete cost effectively with similar products made using other materials such as metals
and plastics. For future research and development, this entails understanding and optimizing
fiber-reinforced cementitious compositions from a fundamental material science perspective;
developing and implementing the use of cost-effective raw materials, particularly reinforcing
fibers and other forms of reinforcement; and developing efficient manufacturing methods
to produce thin reinforced cementitious products.

Ashish Dubey
Editor
Chair, ACJ Committee 549
Thin Reinforced Cementitious Products and Ferrocement

iii
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

iv
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

TABLE OF CONTENTS

Preface .................................................................................................................................. iii

SP-224-1: GFRC-30 Years of High Fiber Cement Composite

. Applications Worldwide ....................................................................................................... 1
by G. T. Gilbert

SP-224-2: Towards Prestressed Thin-Sheet Glass Concrete Products ............................ 21

by G. Vilkner and C. Meyer

SP-224-3: Textile Reinforced Concrete: Investigations at Different Levels ................. 33

by J. Hegger, A. Sherif, 0. Bruckermann, and M. Konrad

SP-224-4: 'Textile Reinforced Concrete (TRC) for Integrated Formworks .................... 45

by W. Brameshuber, M. Koster, J. Hegger, S. Voss, T. Gries, M. Barle, H.-W. Reinhardt, and
M. Kruger

SP-224-5: Exterior Cladding Panels as an Application of

Textile Reinforced Concrete .............................................................................................. 55
by J. Hegger, H. Schneider, A. Sherif, M. Molter, and S. Voss

SP-224-6: Ultra-High Performance Concrete with Ductility: Design, Prototyping, and

Manufacturing of Panels and Boxes .................................................................................. 71
by D. Zakariasen and V. Perry

SP-224-7: New Cement Composites for Thin Structural Products ................................ 89

.by E. Parant and P. Rossi

SP-224-8: Structural Evaluation of Cement Skin Sandwich Building System ........... 101
by Y. Shao, E. Blain-Cosgrove, and B. Robinson

SP-224-9: Properties of Short Fiber Reinforced Cement Paste for Concrete Tubes
Produced by Centrifugation Method ................................... ,........................................... 113
by D. Hesselbarth and J. Kaufmann
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

SP-224---1 0: Temperature Controlled Microwave Acclerated Curing of Precast

Ferrocement Secondary Roofing Slabs ............................................................................ 127
by K. C. G. Ong, C. P. Teo, C. H. Shum, L. H. J. Wong, S. T. Tan, and C. T. Tam

SP-224-11: Freeze-Thaw Durability of Commercial Fiber-Reinforced

Cement Board ................................................................................................................... 145
by K. G. Kuder and S. P. Shah

v
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=University of Texas Revised Sub Account/5620001114, User=erur, ert
No reproduction or networking permitted without license from IHS Not for Resale, 01/26/2015 02:27:05 MST
Daneshlink.com
daneshlink.com

SP-224-12: Crack Growth Resistance of Thin Mortar Layers with Hybrid Fiber
Reinforcement ................................................................................................................... 161
by L. Sorelli, N. Banthia, and G. A. Plizzari

SP-224-13: Modeling of Cement Based Composite Laminates .................................. 179

by B. Mobasher

SP-224-14: Effect of Crack Growth on Overall Mechanical Properties of Cement

Composites ....................................................................................................................... 193
by M. Boulfiza and N. Banthia

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

vi
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=University of Texas Revised Sub Account/5620001114, User=erur, ert
No reproduction or networking permitted without license from IHS Not for Resale, 01/26/2015 02:27:05 MST
Daneshlink.com
daneshlink.com

SP-224-01

GFRC-30 Years of High Fiber Cement

Composite Applications Worldwide

by G. T. Gilbert

Synopsis: Thin, fiber reinforced cementitious products offer a useful balance of

properties such as strength, toughness, environmental durability, moisture resistance,
dimensional stability, fire resistance, aesthetics and ease of handling and installation.

For more than 30 years, AR glass fibers have been at the forefront in the development of
new applications of such products throughout the World. Glass Fiber Reinforced
Concrete [GFRC] is a thin, cement composite based on AR glass fibers with an excellent
strength to weight ratio.

Extensive early laboratory work produced a test method for determining long term
strength. The validity of this work has been proven by the large number of buildings clad
with GFRC, as well as a vast range of other GFRC products, used over a this 30 year
period.

This paper explains the fundamental principles behind GFRC and gives examples of
some of its uses. These applications range from high quality, architectural wall panels
and decorative elements through to modular buildings down to low cost channel sections
and utility components. New developments and techniques will also be discussed.
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Keywords: AR glass fibers; AR mats; AR meshes; decorative elements;

durability; engineering properties; fiber cement; fiber reinforcement;
manufacturing methods; modular buildings; thin cementitious products;
wall panels

1
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=University of Texas Revised Sub Account/5620001114, User=erur, ert
No reproduction or networking permitted without license from IHS Not for Resale, 01/26/2015 02:27:05 MST
Daneshlink.com
daneshlink.com

2 Gilbert
Graham T Gilbert, CChem. MRSC is a Chartered Chemist and a Member of the Royal
Society of Chemistry. He is currently with Vetrotex Cem-FIL, a subsidiary company of
the Saint-Gobain Group, one of the World's leading suppliers ofbuilding materials.

Based in the UK, he has worked on the use of AR Glass fibers as a fibrous reinforcement
for cement based composites and concrete for more than 30 years. During this period, he
has traveled extensively encompassing all the major European countries as well as the
Middle East and North America providing techno-commercial support to both new and
existing manufacturers.

He is also a member of the American Concrete Institute, The Precast/Prestressed Institute

and the International Glassfibre Reinforced Concrete Association.

1.0 INTRODUCTION

Cement based materials have inherent defects such as flaws in the matrix due to shrinkage
and debonding at interfaces. AR glass fibers in this brittle cementitious materials help to
enhance the composite toughness and tensile strength by synergistically interacting with the
micro cracks that develop when the composite is loaded. The AR glass fibers restrain crack
opening and crack growth by effectively bridging across the micro cracks.

The most common form of glass fibers, E-glass, is used as a reinforcing material in resin
composites referred to as FRP. However, when E-glass fibers are exposed to portland
cement based mixtures, such as mortars or regular concrete, the alkaline nature of the
cementitious mixtures rapidly deteriorates the glass fiber. Because of this, AR glass
fibers were developed by intrinsically modi tying the chemical composition of the glass
fibers such that they are inherently more chemically resistant to the alkaline nature of
cementitious matrix.

The actual dose rate used will determine the final composite properties with 0.6kglcu.m.
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

[llb/cu.yd] being used for plastic shrinkage cracking of regular concrete. However, thin
section 12mm [1/2''] GFR Concrete wi11 use much higher dose rates ranging from 2-5% of
total weight being the normal range. Recently, dose rates as high as 15-18% of continuous
filament have been used for newly developed filament wound poles. The exceptional high
tensile strength of AR glass fibers imparts tensile properties to the resultant composite as
well as improving its toughness and impact strength.

Because GFRC has both good tensile and compressive strength as well as being lightweight
with good fire properties and low maintenance, it has been used throughout the World in a
wide range of applications and these will be highlighted later in this paper.

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 3

2.0 AR GLASS FIBERS

2.1 General Data and Fiber Types

In order to make glass fibers resistant to the lime generated during the setting of Portland
cement, zirconium is added to the glass mixture composition prior to melting and fiberising
the raw materials. The added zirconium becomes part of the glass fiber molecular structure
in the manufacturing process and is not just a protective coating. The minimum zirconium
content in the composition for good durability is about 16% by weight. The glass fibers with
this zirconium modification are usually referred to as alkali-resistant glass fibers or AR glass
fibers. AR glass fibers are chemically stable resisting both alkali and acid conditions.
Chemical composition of the AR glass fibers is shown in Table I and the physical and
mechanical properties in Table 2.

AR glass fibers for use in concrete are available in three basic forms - discrete chopped
strands [CSJ, continuous rovings and meshes.

2.1.1 AR Glass Fiber Discrete Chopped Strands-- are used primarily in premix glass fiber
reinforced concrete [high dose rate] and in crack control of concrete [low dose rate] where
the glass is added directly to the cement or concrete slurry. Typically, AR glass fiber CS are
available in two types, integral and water dispersible. Glass fiber CS are made up of bundles
of individual filaments with the typical diameter of these filaments being I2-20 microns.

2.I.2 Integral Chopped Strands-- are designed to stay as bundles of filaments through mixing
and placing, with as little breakdown of the bundle as possible [Fig lA). Integral strand
bundles can contain as many 400 and as 'few as 50 filaments. The number of filaments per
bundle is usually referred as strand geometry. The diameter of the individual filaments, the
number of filaments that are bundled together, and the integrity of the bundle are the key
factors that determine performance characteristics of the strand. The typical length of
discrete AR glass fiber strands used in thin-reinforced products ranges between 6 [I /4"] to 40
[I %''] mm. The strand geometry, strand length, and glass fiber content all contribute to the
processing characteristics ofthe composite and its final properties.

2.1.3 Water dispersible Chopped Strands-- are designed to disperse quickly into individual
strands on contact with water or an aqueous cementitious mixture. These fibers are used in
composites where a fine dispersion of individual mono filaments is desired rather than intact
fiber bundles. In particular, water dispersible AR glass fibers are commonly used to reduce
cracking in concrete, mortars and stucco application and in manufacturing processes that
involve cementitious slurries with initial high water content such as modified Hatschek
process or for calcium and sodium silicate applications frequently using the filter-press
processes. Typical length of water dispersible AR glass fiber strands used in thin-reinforced
cement products ranges between 6 [I /4"] to 25 [I"] mm.

2.1.4 Continuous AR Glass Fibers-- are available in the form of roving [Fig. IB]. A
roving is an assemblage of several continuous AR glass fiber monofilaments. The

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

4 Gilbert
manner in which the monofilaments are assembled varies and differentiates one roving
type from another. Fundamentally, the construction of an AR fiber roving is as follows:
• Several continuous alkali-resistant glass fiber monofilaments are gathered
together to form a continuous strand. The typical diameter of the individual
alkali-resistant glass fiber mono filaments ranges between I 0 to 20 microns.
Typically, the number of monofilaments that are gathered together to form a
continuous strand ranges between 50 to 400.
• Several continuous strands as explained above are assembled together to form a
continuous roving. Typically, the number of continuous strands that are
assembled to form a continuous roving ranges between 20 to100.

2. I .6 Glass Fiber Meshes and Mats-- are woven or dipped, non-woven products from
assembled glass fiber rovings or strands whilst mats are chopped fibers bonded together
with a polymeric coating. Traditionally they were woven products of heavily coated E-
glass fiber yam used mainly in the production of cement boards (Venta et al. I995, I997,
I 998). As an alternative, an alkali-resistant (AR) glass fiber mesh can be used to reduce the
need for the coating. The AR mesh has recently been used in a new system for seismic
improvement of masonry walls.

3.0 PRODUCTION METHODS

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
3.I Simultaneous Spray Process

In the simultaneous spray process, continuous AR glass fibers are chopped through a gun
and air-sprayed simultaneously with the cement slurry onto a mold surface, [Figure 2].
To spray the entire mold area, both the fiber and cement slurry spray guns are moved to
follow the contours of the mold. Successive 4-6mm [~114"] layers are sprayed and roller
compacted to form a typical 12-15 mm thick (~w·] panel. The compaction removes air,
helps bond and ensure a good quality finish.
The simultaneous spray process can be manual or automated. Spray process allows
tremendous flexibility in manufacturing complex architectural shapes as well as
producing a high strength product. Consequently, architects around the globe commonly
design and specify architectural shapes manufactured using the spray process.
3 .2 Premix Process

The premix process consists of first mixing together the other ingredients (cement,
sand, admixtures, water, etc.) in a standard or specialized mixer to give a low viscosity
mortar. The AR chopped strands are added to this mortar and cast to from a thin product of
desired shape in a mold. The casting process may be similar to concrete casting or hand
packing [lower strengths] or it may not involve spraying of fiber-cement slurry as a method
to fill the mold and/or vibration to achieve satisfactory slurry compaction in the mold. In the
premix process, the maximum amount of fibers that can be incorporated in the mixture is
dependent upon the length and diameter of the fibers used. Additives such as polymers,
pozzolans e.g. metakaolin, and/or flow aids such as water reducing agents are generally used
to facilitate the mixing operation. The premix process typically yields a three-dimensional

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 5

random orientation of fibers in the mixture. Consequently, premix products are not as
strong as simultaneous sprayed ones but the process has the advantage of ease of
production and a lower level of skill required to produce the end product.

3.3 Filament Winding Process

Filament winding process was developed for FRP composites but has recently been used
to produce GFRC poles.
Figure 3 shows a pole being wound onto a mandrel with AR glass fibers
The fiber roving strand passes over several round steel bars placed below the level of a
specially modified mortar mix before being wound onto a mandrel.
Various continuous fiber cement based composites consisting ofunidirectional lamina,
cross ply and angle ply laminates can be manufactured with a typical AR glass volume
fraction ofl3%. Mechanical properties of wound tube test specimens gave tensile
strengths higher than 300 MPa (50,000psi] and flexural strengths as high as 200 MPa [30,
OOOpsi]

The filament winding method has been refined commercially in the US for the production of
a range of poles and inductively transparent, high temperature ladles with glass fiber volume
fraction ranging from 10 to 25%. [Ref US Patents 5039345 (Mott 1991) and 5880404
(Stanley and Mott 1999).

3.4 Filter-press Process

In the filter-press process, first a wet fibrous mix is produced with an excess of water. Then
this mix is charged into a mold, which has a perforated plate at the base. A filter material is
laid on top of the mold base. The mix is then pressed by a top plate, which squeezes out the
excess water through the base of the mold and through a small gap between the top plate and
the sides of the mold. The compressed board or tile is then removed from the mold and
stacked for curing. Depending on the shape, the product can be demolded immediately
whilst in the unhardened state. It is also possible to use rapid setting cements to
accomplish instant demolding. The filter-press process is well suited for mass production
of products having simple or complex shapes.

4.0 GLASS FIBER REINFORCED PRODUCTS PROPERTIES

The physical and mechanical properties of glass fiber reinforced concrete are discussed more
fully in ACI 544.1 R-96. The following summarizes these properties.

The mechanical properties of GFRC composites depend upon the fiber content,
water/cement ratio, density, sand content, fiber orientation, fiber length, and polymer content
if used. Typical properties for traditional spray up GFRC containing 5% by weight of glass
fibers are shown in Table 3 (PCI MNL-128-01). As shown in this table, GFRC composites
have significant load and strain capacity. Whilst some of these properties reduce a little with
time for standard GFRC their real time performance is well documented and all accepted
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

6 Gilbert
design procedures allow for it in establishing design values. PCI publication MNL-128-01,
'Recommended Practice for Glass Fiber Reinforced Concrete Panels' details the most widely
accepted design procedure in the industry.

If reduction in mechanical performance of composites is of concern, GFRC composition can

be modified in several ways to prevent this from occurring. The formation of calcium
hydroxide within the fiber strands has been held to be largely responsible for the change in
properties with time (Bentur 1985). The measures that are used to arrest the change in
properties generally attempt to prevent the formation of calcium hydroxide. The glass fiber
manufacturers have made available alkali-resistant glass fibers with special coatings that
reduce the affinity of the fibers for calcium hydroxide (Hayashi, Sato, and Fuji 1985). Most
other methods to improve durability of GFRC rely on either the use of pozzolanic
admixtures such as silica fume, metakaolin or fly ash, (Marikunte, Aldea, and Shah 1997;
Soukatchoff 1999; Soukatchoff and Ridd 1991; Purnell et al. 1991) or use of special
cements such as calcium sulphoaluminate cements that do not produce calcium hydroxide as
a hydration product (Molloy, Jones, and Harmon 1993; Molloy and Jones 1993). Acrylic
thermoplastic co-polymers have also been reported to reduce the extent of reduction in
mechanical performance with time (Ball and Wackers 2001 ). Acrylic thermoplastic co-
polymers are usually used in GFRC products because they improve the curing of the GFRC.

In addition to traditional spray up GFRC, premix glass fiber reinforced concrete is

growing in use for certain products. Typical properties for premix GFRC are shown in Table
4 (PCI MNL-128-01). Generally premix GFRC has a lower fiber content, uses shorter
fibers, and has significantly greater three dimensional fiber orientation than the largely two-
dimensional orientation obtained with spray up GFRC, which all contribute to it having
lower mechanical performance than the spray up GFRC.

5.0 APPLICATIONS OF GFRC

5.1 Cladding

5.1.1 Modular Buildings, -- Single- or two-storey modular buildings, have been

constructed with AR glass fiber reinforced cementitious sandwich panels integrated into a
lightweight steel structural frame during erection. Sandwich construction of the panel
involves two 8 mm [::::3/8in] thick AR fiber reinforced concrete skins attached onto both
sides of a !55 mm [6in] thick core of lightweight concrete. This building system has been
independently tested for load capacity, sound insulation, thermal conductivity, and fire
resistance.

Figure 4 & 5 shows an example of such a system.

Much earlier, in the late 1970's, GFRC panels were used on exterior wall of prefabricated
timber frame houses constructed to meet the shortage of dwellings in Scotland. The
panels were typically 10 mm [0.4in] thick and had an aggregate finish surface. Simple
cast-in washers for face fixing the panels were incorporated at 50mm [2in] on center.

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 7

Wind pull off tests conducted on the wall systems yielded results in excess of those
needed for the 200 kmlhour (125 mph) gusts found in that region. These houses were
independently inspected by the GRCA (Glass Reinforced Concrete Association, UK)
after 20 years and were found to be in good and serviceable condition. Figure 6 shows
these dwellings after 20 years exposure in the Highlands of Scotland.

5.1.2 GFRC Architectural Facade Panels -these panels can be manufactured as wall and
window units, spandrel, soffit and fascia panels, mansard roof elements as well as
mullions, cornices and column covers. Figure 7 shows a photograph of the Cervantes
Convention Center situated in St. Louis, Missouri in which 1670 m2 [:::::18000ft2 ] of the
building exterior was clad with GFRC architectural faryade panels. The panel size varied
but the average was approx. 2.4 m x 6.0 m. [:::::8 x 20ft] The panel skin consisted of 12.5
mm [l/2in] thick GFRC plus a 6 mm [1/4in] thick facing mix attached to a structural
steel frame, which in tum was attached to the building. In several panels, two finishes
were combined on the same panel. Brick red finish on the panels was achieved with the
use of white cements, sands and pigments. The intricate architectural details on these
panels were created by forming the panels over rubber liner molds. Some panels also had
limestone finish. This was achieved with the use crushed stone, pigment and
sandblasting. When using facing mixes, care is required to reduce incompatibility from
thermal and moisture movement.

There are innumerable examples of high quality, architectural panels on buildings

completed for over 30 years around the World. Fig 8 shows the recently completed Nile
City Project in Cairo and Fig. 9 the San Francisco Towers complex in California, USA
where 155 000 GFRC panels were used.

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
5.1.3 GFRC Decorative Elements- GFRC is capable of closely imitating natural
materials, which tend to be expensive and in short supply. Thin complex shapes with
excellent surface finish and surface details can be easily formed using AR glass fiber
reinforced cementitious materials. Molds to form these shapes are frequently taken from
deteriorated original carvings. GFRC architectural elements are relatively light in weight
and require low maintenance. These attributes make GFRC architectural elements a
sensible choice for both new and refurbished buildings. Figure I Oa & b and Fig. II show
some applications.

5.2 Road and Rail Sound Walls

Throughout the world, new highways and mass transit rail systems compete for space in
already developed urban areas. The result is that major traffic routes are found closer to
commercial and residential areas and it becomes necessary to suppress noise pollution to
the surroundings. GFRC noise barriers are being increasingly used since they are light
in weight and offer simplicity and speed of erection without requiring the use of heavy
lifting machinery. This gives reduced disruption to traffic and greatly reduced loads on
any elevated structures allowing for the same material to be used throughout. The
moldability of GFRC allows for aesthetically pleasing designs which are more attractive
and acceptable to both residents and travelers as well as to bridge engineers and

Copyright American Concrete Institute

8 Gilbert
architects alike. In addition, GFRC has excellent resistance to salt attack, freeze-thaw
and rotting thus reducing maintenance.

Figure 12 shows a GFRC reflective sound wall in Spain and Fig 13 a different design,
absorbing, in South East Asia.

5.3. Ducts and Channels

5.3.1 Drainage Channels- GFRC has been used for drainage and transporting liquids
represent another application for GFRC. Fig 14 shows a commercially available high
volume, rain-water drainage channel used in parking lots, road and highway applications.
These channels are designed for optimum flow capacity and are available in different
cross-sectional sizes with lengths ranging up to 2 meters (6.6 feet). Further, these
channels are lightweight, easy to install in long sections with reduced excavation,
maintenance free, and require fewer silt traps or manholes due to their superior hydraulic
performance. The channels are produced by vibration casting an AR fibers mix into a
two-part mold.

5.4.2 Cable and Pipe Ducts and Conduits in GFRC are used in Europe, Japan and the
USA. Figure 15 shows GFRC pipe trench liner in USA.

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
5.5 Tunnel and sewer Linings

5.5.1 Tunnel Liners --GFRC has been widely used for tunnel lining applications and
Figure 16 & 17 shows Heathrow Express railway station at Heathrow Airport, London,
UK. The panels used in this application are nominally 12 to 18 mm ('h" to%") thick and
are made from a cementitious mixture reinforced with alkali-resistant glass. Lining
within the stations consisted of 9000 acid-etched and sand-blasted panels of size 1.8 m
(5.9') wide, 0.90 m (2.95') wide and 0.70 m (2.3') deep. A 50 mm (2") deep recess in
each panel allowed enamelled glass advertising panels to be secured within.

5.5.2 Sewer Linings- -Figure 18. shows a photograph of thin reinforced cementitious
sewer lining application. GFRC sections are very versatile and cost effective, and
thereby offer several significant advantages over FRP in sewer applications. They are
stiffer and bond well with the grout lining, essentially becoming part of the sewer
structure. Consequently, they are more resistant to the damaging influences of water
pressure or ground movement.

5.5.3 Canal Bank Protection Linings in GFRC is demonstrated in Figure 19. Such linings
are used to prevent erosion of the canal banks caused by different sources such as
hydraulic discharge and incidental contact with passing boats. The panels are typically 6
to 9 mm thick, with ribs on rear face for strength and stiffness. The typical size of the
panels is about 2m x 1.36 m (6.5' x 4.5').

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 9

5.6 Pipes and Poles

Filament wound GFRC poles have recently been developed in North America. Figure 20
shows a pole manufactured using continuous alkali-resistant glass fibers. These poles
can have a fiber volume fraction as high as 25%. In terms of their mechanical behavior,
these products are exceptionally strong in tension (notched tensile strength of about 90
MPa), compression (compressive strength of about 175 MPa) and flexure (bending
strengths of about 150 MPa). Using the filament winding process, the pole products can
be easily manufactured in lengths of up to 15 meters (50 feet). Potential applications of
the pole products include induction and wireless transmission-invisible poles and
pem1anent formwork for seismic and marine columns.

5.7 Bridge Parapets

Fig. 21 shows GFRC panels used on the BTS Skytrain, Thailand's first mass transit
system. The project extended about 23 km (14 miles) through the heart of the Bangkok
and was entirely elevated. The parapets shown in this figure are typically 1.1 m high x
2.7 m [3' 6" x 8' 9"] long, and nominally 15 mm [2/3"] thick with ribbed top and bottom
using expanded polystyrene void-formers. The ribbing provides the panels with
additional strength and rigidity to resist the high wind loads caused by the passing trains.
The parapet panels shown in this figure are reinforced with discrete, alkali-resistant glass
fibers.

5.8 Landscaping Products

Glass fiber reinforced concrete is used widely for simulated rock installations in zoos,
hotel and office lobbies, swimming pools, climbing walls, golf courses, and theme parks.
Thin cementitious panels are usually factory prefabricated using the spray up process or
sprayed premix. The molds are usually made of rubber that has been cast off an actual
rock face. The cast panels are then assembled on the job site. The panels can be
integrally colored or can be colored on the job site using acrylic stains.

6.0SUMMARY

Glass fiber reinforced concrete, GFRC is a thin form of concrete with the high tensile
strength of the AR glass fibers complimenting the high compressive strength of the
cementitous matrix. The resulting composite, usually only 12-25mm [lh -1 "] thick, offers
a unique balance of properties such as strength, toughness, dimensional stability,
environmental durability, moisture resistance, freeze thaw resistance, fire resistance,
esthetics, and ease of handling and installation.

Because of this blend of attractive properties, GFRC has found a wide range of uses in more
than 40 countries spanning a period of more than 30 years.

The fibers are easy to incorporate using a variety of production methods to suit the end need.

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

10 Gilbert
Applications range from the highly visible large architectural panels on low and high rise
buildings to decorative elements to more mundane uses such as ducts and channels and
tunnel linings.

Table !-Chemical composition ofAR-glassfibers, percent by weight (PC/ MNL-128-01)

Component AR-elass
Si02 61.0-62.0
Na20 14.8-15.0
K20 0.0-2.0
AfzOJ 0.0-0.8
Zr02 16.7-20.0
Ti02 0.0-0.1
Li 20 0.0-1.0

Table 2-Properties AR-glassfibers (PC/ MNL-128-01)

Property AR-Giass
S_j)_ecific Gravity 2.70-2.74
Tensile Strength, MPa [psi] ] 700 [2.5 X I 051
Tt
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Modulus of Elasticity, GPa [psi] 72 0.4 X 106-]

Strain at Break, % 2.0

T•ble 3: TJ·pic•l range of tr•dilionul sprayed GFRC properties' (PCI MNI.-128-

01)

Propertv 28·dav Al!ed

Drv Densitv Kldcu.m. loctl 1900-21001120to 1401 1900-2100 1120 to 140i
Compressive MPa [psi] 50-82 17·12,000] 70-82 fl().> 12,0001
Flexural: MPa [psi]
Yield(FY) 6-10 [900 ->1,500] 7·11 (1.000 to 1.600]
Ultimate strength (FUl 14-24(2000 to 3,500j 9·17[1300 to 2,500j
Modulus of elasticitY GPa [psi I 7-2111->3.0xl06 l 18-2SI2.5-> 4.0xJo•l
Direct Tension: MPa [psi]
YicJd(n·r S- 7 [700 to 1.000) 5·7 [700 to 1,000)
Ultimate strength (Til) 7-11 [1.000 to 1,600] 5-8 [725 to 1,100)
Strain to failure % 0.6 to 1.2 0.03to 0.08
Shear: MPA [psi]
lntcrlaminar 3-6 [400 to 800] 3-6 [400 to 800)
In-plane 7-11· (I 000 to I 600] S-8 f725 to I I001
Coef. of thermal expansion f"C 1o-2o x w·• 10·20 x 10"'
(in.iin.1"FI 1"'12xl0'6 l 1=12x10.. l
Thermal conductivity. W-'m C 0.5 to 1.0 O.Sto 1.0
tBtuiin.'lulft'I"Fl [3.5 to 7.0] 13.5 to 7.01

Note:
'These are typical values and are not to be used for design or control purposes. Each
manufacturer must test production composites to establish properties for design. The
values achieved in practice will be dependent on mix design, quality control of
materials, fabrication process and curing. Cement/sand ratio in the above composites
ranges between I : I to 3: I.

'Developed from accelerated testing progroms on GFRC specimens immersed in 50 to

SO"C ( 122 and 175'f) water. On the basis of comparisons between behavior in real
weather and accelerated tests, predictions can be made of properties for 50+ years in
different climates.

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 11

Table 4: Typical ran2e of premix GFRC properties• (PCI MNL-128-01)
Prooertv 28 day
Metric Equivalent
Densitv (drY) 1800-2000 kf{lcu.m 110 to 130 pcf
Compressive Strenlrth 40-60MPa 6,000 to 9,000 psi
Flexural:
Yield (FY) 50-BOMPa 700 to 1,200 psi
Ultimate strength (FU) 10-14MPa I ,450 to 2,000 psi
Modulus of elasticity 7-20GPa l.Ox1 06 to 2.9x I06psi
Direct Tension: Yield (TY) 4-6Mpa 600 to 900 psi
Ultimate strength (TU) 4-7MPa 600 to 1,000 psi
Strain to failure 0.1-2% 0.1 to 0.2 oercent
Shear: ln-olane 4-7MPa 600 to I 000 psi
Coefficient of thermal exoansion. 10-20x 10.()7c Aoorox. 12x I O.f! (in./in.I"F)
0.5 to 1.0
TI1ennal conductivity 3.5 to 7.0 (Btu/in./hr/ft2j0f)
Wlm°C
Note: "These are typical values and are not to be used for destgn or control purposes.
Each manufacturer must test production composites to establish properties for design.
The values achieved in practice will be dependent on mix design, quality control of
materials, fabrication process and curing.

Fig lA AR Glass Fiber Chopped Strands

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

12 Gilbert

Fig I B AR Glass Fiber Continuous Strand

Roving

Fig I CAR Glass Fiber Mesh

Fig 2 Simultaneous Spraying of GFRC

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 13

Fig 3 Filament Winding of a pole with AR Glass Fibers

Fig 4 GFRC Insulated Wall Panels for Modular Housing in Dubai

V'!~?'r
.A•u.4l~.os
•b.:.s~-

l h•lt'·e_<'t~l~if'>A t\l«<t'H Ct<>t,~~ tN-1.f'l

l\ad~~~ ~h-N.b Rvl:':fl~ ::o~~·fl:' ;t~>Q,.~:tu l\t:dt~•<.>m 1 kdt~«=St :;. ~lhf~lfl:l

l;~lh*:t" o.:.>rn e '~""'=' ~dm>.•~ l ~tbruoll' to~.lt.:~,- hm~t t~nJ(~

Fig 5 Floor Plan of Modular Construction House

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

14 Gilbert

Fig 6 Timber Frame Houses with Single Skin GFRC Panels. Built 1976, Scotland

Fig 7 1670 sq.m of GFRC panels on the Cervantes Convention Centre,

St Louis, Missouri. 2 finishes - Ornate/pigmented and stone simulation
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 1~

Fig 8 Nile City project, Cairo,Egypt. 2 x twin Towers

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Fig 9 155 000 GFRC panels on San Francisco Towers, California, USA

Copyright American Concrete Institute

16 Gilbert

Fig 1Oa Columns, Arches and Domes With Surface Details

Fig 1Ob 1.3m [5ft] Capital Unit -View from Rear

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 17

Fig II Terra Cotta Replacement on Shepard Hall, New York 72000 units with more than
4000 shapes inc. sculptures
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Fig 13 Absorbing Barrier S.E Asia

Copyright American Concrete Institute

18 Gilbert
.·· ·.. ·: ... ·. .
.........
·~<~::-.·· .
, ••· · - - . . • ¥

... -····
, ...., .....
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Fig 14 Engineered, Optimum Flow, Drainage Channels, UK

Fig 15 Pipe Trench Liners, USA

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 19

Fig I 6 & I 7 Heathrow, London

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Fig I 8 Sewer Re-lining

Fig I 9 River Bank Protection, Holland

Copyright American Concrete Institute

20 Gilbert

Fig 20 GFRC Pole 7.5m [25'], USA

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Fig 21 Bridge Parapets, Thailand

Copyright American Concrete Institute

SP-224-02

Towards Prestressed Thin-Sheet Glass

Concrete Products

by G. Vilkner and C. Meyer

Synopsis: Thin sheet concrete products are receiving increased attention because of the
large number of potential applications. By using crushed glass as aggregate, a multitude
of different esthetic effects can be produced, which again open up numerous architectural
and decorative uses. Such thin sheets are most effectively reinforced with fiber mesh,
whether made of polypropylene, AR-glass, or other types of materials.

At Columbia University, a project is currently under way to explore the possibilities of

prestressing thin sheet glass concrete products. There are numerous performance criteria
that need to be satisfied by the fiber mesh material in order to qualifY for the tasks on
hand. Most promising to date are high-performance materials such as aramid and carbon
tiber mesh.

This paper discusses the elimination process by which the most appropriate type of fiber
mesh was selected. Various technical problems of prestressing and anchoring the fiber
mesh are pointed out, as well as other issues that need to be resolved, before such
products can be mass-produced commercially.

Keywords: aramid; fiber -reinforced concrete; glass concrete; prestressed

concrete; textile reinforcement; thin sheets

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute 21

22 Vilkner and Meyer

ACI member Gregor Vilkner works with LZA Technology, a division of The Thomton-
Tomasetti Group in New York City. He obtained his Ph.D. in Civil Engineering from
Columbia University in the City of New York and also holds a Dipi.-Ing. from the
University of Rostock in Germany. His research interests include high-performance
polymeric fibers, instrumentation engineering, and concrete technology.
Christian Meyer, FACI, is a Professor of Civil Engineering at Columbia University in
the City of New York. He is a member of ACI Committees 446, Fracture Mechanics;
544, Fiber Reinforced Concrete; 555 Concrete with Recycled Materials (Chair); Joint
ACI-ASCE Committee 447, Finite Element Analysis of Reinforced Concrete Structures;
and the ACI Board Task Group on Sustainable Development.

INTRODUCTION
Thin-sheet concrete products have attracted the attention of researchers and concrete
producers alike in recent years because of their numerous potential applications. In
conventional steel reinforced concrete elements, the cover needed to protect the steel
against corrosion calls for a minimum sheet thickness of at least 5 to 7 em. The tendency
of the ribs of standard reinforcing bars to spall off thin concrete covers may require a
further increase of the minimum plate thickness. For non-metallic reinforcement no
corrosion protection is needed, and thicknesses of a few mm are theoretically possible.
Woven fabrics or fiber mesh, also referred to as textile reinforcement, have proven to be
a viable form of such reinforcement.· The rovings are curved at points of intersection,
caused by the weaving process. It has been observed by other researchers that woven
fabrics, when stressed as ordinary reinforcement, need to be straightened before they
contribute in the load carrying process (Curbach 1999). This delay inhibits distributed
cracking to some extent, but if the fabrics are stretched slightly before being built in, such
curvature effects become negligible. Prestressing the embedded reinforcement, whether
provided in the form of single rovings or continuous fiber mesh, further improves the
mechanical properties of structural members and enhances their durability because of the
absence of cracks (Kriiger 2004, Vilkner 2003).
The substitution of crushed glass for natural aggregate opens up additional options,
primarily in the field of architectural concrete, because of the esthetic potential of colored
glass. An important prerequisite is an effective measure to counter the potentially
damaging effects of alkali-silica reaction (ASR). At Columbia University, a major
research project has been under way for a number of years to utilize waste glass as
aggregate in concrete products, and the work reported in this paper is a part of this
ongoing larger effort. Some of the commercially available fiber mesh materials shall be
described and compared and the various properties of prestressed thin sheets discussed.
This paper describes work in progress, pointing out some of the issues of mechanical
behavior involved and technical problems that need to be overcome, before such thin
sheets can be mass-produced commercially.

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 23

PERFORMANCE SPECIFICATIONS FOR FIBER MESH
Fibers have been used in the concrete industry for some time to improve the tensile and
flexural performance of concrete (Balaguru and Shah I 992). Also the beneficial effects of
such fibers on shrinkage cracking and impact and fatigue resistance are well documented
( Mindess et al. 1987, Ramakrishnan et al. 1995). While short randomly distributed fibers
made of hydrophobic synthetic materials have some advantage during the mixing
process, their relatively poor bond properties, combined with their random orientation in
three dimensions, make their performance less than optimal. With continuous fiber mesh,
some of these disadvantages are eliminated. Their use as reinforcement of thin
cementitious sheets commenced in the mid- I 980s (Gardiner and Currie 1983, Daniel and
Shah 1990, Peled et al. 2000). In particular, the bond properties of textiles have been
studied in detail in recent years (Bentur et al. 1997, Peled et al. 1998). When considering
ravings instead of single fibers, as used in textiles, it is essential to address the question,
if only the outer fibers of a roving develop bond to the matrix and how deep the matrix
material penetrates into the rovings. Also the influence of the type of fabric, e.g., woven
vs. knitted, is a major point of interest.
In order to select a fiber mesh material that is suitable for prestressing thin sheet glass

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
concrete products, it is necessary to evaluate each potential candidate using the various
performance specifications for the different applications. Foremost among such
requirements is the chemical stability of the material in the alkaline environment of the
cement paste. For prestressed concrete to achieve its traditional advantages, the
reinforcing material has to be of high strength and show little stress relaxation at the
intended prestress level. Related to this requirement is that of a reasonably high static
fatigue strength. The melting point needs to be high enough as to cause no significant
creep effects at elevated service temperatures, which would lead to an unacceptable
lowering of the effective prestress. Good bond properties are required, in particular in the
end zones, where otherwise separate anchorage mechanisms would be needed. Ideally,
the material should also possess good ductility and have a large energy absorption
capacity to assure gradual flexural failure modes.

FIBER MATERIAL SELECTION

All fibers are made of either inorganic or organic material. The inorganic category
includes materials such as metals, minerals, ceramics, carbon and glass. The list of
organic fiber materials, on the other hand, seems to be limited only by the creativity of
nature and the chemical industry. Nature still holds the record for the strongest fibers,
which are spun by spiders. Other natural fibers include cellulose, silk, and cotton. Man-
made organic fibers include nylon, polypropylene, polyvinyl alcohol (PVA),
polyethylene, and aramid, among many others.
High-strength steels, which traditionally have been the materials exclusively used for
prestressed concrete applications, need to be protected against corrosion. Stainless steels
are available, including piano wires. These are relatively expensive and are not being
offered in the form of woven or knitted meshes. Therefore, they shall not be considered
here any further. Most of the commonly used polymeric. materials, such as

Copyright American Concrete Institute

24 Vilkner and Meyer

polypropylene, nylon, and polyvinyl alcohol, have relatively low melting points and
exhibit excessive creep and relaxation behavior, which makes these materials unsuitable
for prestress applications. Alkali-resistant (AR) glass, aramid, and carbon are more
suitable in those regards. Because of their very high unit strengths, relatively low creep
deformations, and high melting points, these materials are now being studied for their
suitability for prestressing thin-sheet glass concrete elements. All three materials have in
common, in contrast to metals, that they exhibit highly linear-elastic brittle behavior, and
they are available in the form of woven or knitted fiber mesh.
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Alkali-resistant glass, although popular as short fiber reinforcement, has recently been
shown to have a rather low static fatigue limit (Reinhardt 2002) and therefore is not likely
to be suitable for applications in which the mesh is to be prestressed. Aramid and carbon
fibers have reasonably high static fatigue limits, relative to their ultimate strengths.
However, accounting for their low ductility requires comparably large safety factors
against failure, such as for instance when glass is employed in structural applications.
Especially in the case of carbon, its very high stiffness leads to very low strain levels, far
below those common in conventional prestressed concrete applications. The costs of
aramid and carbon fiber mesh are relatively high for typical thin-sheet applications,
compared with some of the other available materials. However, manufacturers are
steadily improving their production technologies. The expected price reductions will
make new applications feasible, and the resulting increase in demand for these materials
is likely to prompt further cost reductions due to the economy of scale. For these reasons,
they are currently being considered for further study. The emphasis in this study was on
aramid fiber mesh.

ARAMID
Aramids are a family of nylons, including high-strength fibers that are known under trade
names like Kevlar and Nomex. Aramid is a short form for aromatic polyamides. Fig. I
illustrates how amid groups connect phenyl rings to form monomers that build polymers.
The aromatic rings differentiate them from non-aromatic polyamides that form fibers like
Nylon 6,6. In the case of Nylon 6,6, the amid group is found in both illustrated forms, the
cis- and trans-configuration. Single polymer chains can only form in a perfectly straight
fashion in the trans-configuration. If the amid group forms in the cis-configuration it
causes the path of the polymer to change. In aramids, the trans-configuration is formed
almost exclusively, which allows for perfectly stretched polymer chains. A valid and
often used analogy is comparing cooked (cis-) with uncooked (trans-) spaghetti. The
small but important difference between para- and meta-amids, i.e. Kevlar and Nomex, is
that in the first case, amid groups are attached to phenyl rings at carbon atoms directly
across from each other, i.e. at positions I and 4, Fig. I. In the second case, amid groups
connect to phenyl rings at the I and 3 positions, which causes the polymer path to bend
slightly. Both aramid forms are very resistant to high temperatures and chemical attack.
The fact that para-aramids stretch out more perfectly allows them to form fibers in which
the polymer chains are packed more closely, which is the reason for their higher strength.
A more detailed description of the fiber production and the deformation mechanisms is
given in (Meyer and Vilkner 2003). DuPont developed the most popular of these fibers

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 25

under the trade names Kevlar 29 and Kevlar 49 in 1966. In Europe, the AKZO Group
fabricated the competing para-aramid fiber Twaron. Some representative technical data
are summarized in Table 1.
Aramid fibers have basically good chemical resistance (Kasperkiewicz and Reinhardt
1992), except in environments with extreme pH values, where hydrolysis can degrade or
literally dissolve fibers. Concerns do exist that the alkalinity of the pore solution in
Portland cement paste can also cause their deterioration, so that the fibers are often
provided with protective coatings such as PVC, polyester, or epoxy (Broadway 2002). It
remains to be investigated whether the admixtures that are added to the cement matrix to
suppress alkali-silica reaction in concrete with glass or other reactive aggregates are
equally effective in preventing the deterioration of aramid fibers embedded in glass
concrete thin sheets.

MECHANICAL BEHAVIOR OF ARAMID MESH

Aramid fibers have cross-sectional diameters of the order of 10 J.lm 1• Several hundred of
them form a roving or thread. The rovings tested in the present investigations, 700 denier
Twaron fibers, failed at a tensile load of about 0.15 kN (35 lbs). Load-deformation curves
for various fiber materials are reproduced in Fig. 2a (Naaman 2000), whereas Fig. 2b
depicts the behavior of an aramid sample consisting of 8 rovings, as obtained in the
present investigation.
Rovings are being woven or knitted into various textile patterns. For our purposes,
orthogonal meshes appear to be most appropriate. The fiber mesh is woven like standard
textile fabrics such that orthogonal sets of rovings are not interconnected at their points of
intersection. If regular mortar is used, the number of rovings per inch can be quite large.
In glass concrete, however, it is desirable to utilize aggregate particle sizes of at least 3
mm (1/8 inch) to achieve certain esthetic effects. The fiber spacing needs to be larger
than the largest aggregate size, so that the concrete layers on both sides of the mesh are
adequately connected. For this reason, the mesh sizes considered in the present
investigation are 5 by 6 rovings per inch.
In welded wire mesh, orthogonal layers of wire are spot welded at each point of
intersection, which provides excellent mechanical bond. In woven or knitted mesh of
polypropylene or nylon fibers, the bond of weft or warp yams is also improved by the
anchorage provided by the fill yams in the orthogonal direction. In the case of high-
strength aramid fiber mesh, that added anchorage effect is all but negligible, and the only
source of anchorage is an adequate development length. If such mesh is to be prestressed,
it is necessary to provide a separate means of anchorage, which allows to stress and
securely hold several layers of fabric. A first trial for such an external anchorage is shown
in Fig. 3. The edges of two layers of mesh were encased in epoxy bars 12 mm (1/2 in)
thick and 25 mm (1 in) wide, which were used to stress the mesh and stay in place as

1
In the fiber industry, it is common to specify fibers in units of tex or denier, which
indicate the weight in gram of a I 000 m or 9000 m long single fiber, respectively, i.e. 9
denier= I tex. --``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

26 Vilkner and Meyer

pennanent anchorage blocks. A commercial medium-strength epoxy mortar with a
conventional filler was used. However, the choice of mortar becomes an issue of more
importance, if more layers of fabric are to be utilized to generate higher prestress levels.
The major problem with casting the scrim fabric into an epoxy end-block is to assure that
all fibers are aligned parallel and are of equal length in their unstressed state. In practice,
this is difficult to achieve. Therefore, in actual tests the ultimate strength (average stress
per fiber at failure) dropped from 2800-31 00 MPa (400-450 ksi), when few rovings were
tested (e.g., 8 rovings, see Fig. 2b) to below 2100 MPa (300 ksi), when a specimen with
34 rovings was considered. The elastic modulus, i.e. the slope of the linear branch of the
stress-strain diagram is in both cases equal at about 110 GPa (16,000 ksi) (Table 2).
The behavior of an aramid sample consisting of 34 rovings is depicted in Fig. 4, both as a
load-versus-time curve and the standard load-defonnation graph. The experiment was
carried out on a mechanically driven INSTRON testing machine in displacement control
without feedback control. Individual rovings are seen to fail before others reach their
ultimate strength. Therefore the average strength decreases with the number of rovings
tested in parallel. The reason is the different amounts of slack within the individual
rovings prior to load application, coupled with the lack of ductility. If the rovings were
made out of steel, their plastic deformations would assure load sharing between highly
and lowly stressed rovings such that all of them would fail at the same time. Because of
their brittleness, aramid fibers do not lend themselves towards such beneficial load
sharing, and as a result, greatly lowered average strengths have to be accepted in addition
to the relatively large factor of safety, which is a standard requirement for designs
involving brittle materials. It remains to be seen whether under such constraints the cost-
to-benefit ratio of high-strength fiber mesh still makes economical sense.

MECHANICAL PROPERTIES OF GLASS CONCRETE

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
Using glass aggregate in a concrete or mortar mix has fundamental effects on both the
fresh and hardened concrete. The practically nonexistent water absorption of glass
particles has a positive effect on the flow properties of the fresh concrete, whereas the
admixtures necessary to suppress ASR have a negative influence on the workability. By
careful mix design, very high-quality glass concrete can be produced. For example,
concrete systems modified with metakaolin have been reported to profit from a superior
Portlandite-free microstructure, which explains its excellent durability, low shrinkage and
substantially lowered creep deformations (Sabir et al. 2001). Since no aggregate larger
than 6 mm (114 inch) is used, the material should be referred to as mortar instead of
concrete, according to the strict definition of the American Concrete Institute. However,
because of the arbitrariness of this definition we shall take the liberty of calling the
material "glass concrete".
Table 3 summarizes the proportions of such a concrete mix, and a typical stress-strain
diagram obtained after 7 days is shown in Fig. 5. Though the w/c ratio, adjusted to
achieve exceptionally high flow, is higher than in other glass concrete applications, the
material has a very high strength of about 35 MPa (5 ksi) after 20 hours and 70 MPa (1 0
ksi) after 7 days. Based on the 20 hour compressive strength an effective prestress of 14
MPa (2 ksi) was targeted for first trials of actually prestressed thin sheets. Glass concrete

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 27

is characterized by near-linear behavior up to about 80% of the ultimate strength. The
Young's modulus is about 22 GPa (3200 ksi).

PRACTICAL CONSIDERATIONS
The selected epoxy endblock system proved to be suitable for investigative purposes, but
could not sustain an effective prestress of 14 MPa (2 ksi). The 34 rovings were
distributed in 2 layers. Selecting 50% of the failure load of 3.5kN (800 lbs) as
prestressing force on a 75 mm x 12 mm (3 in x 112 in) cross section, the corresponding
prestress was only 1.8 MPa (250 psi). The next step taken was the development of a 7
mm x 25 mm (1/4 in x 1 in) endblock containing 7 layers of fabric, totaling 35 rovings
(Meyer and Vi Ikner 2003). Ultimately, larger rovings had to be considered.
Structural engineers familiar with prestressed concrete are used to deal with high-strength
steels that undergo considerable plastic deformation before failure. Also, stress averaging
is a common concept in structural engineering and justifies the neglect of many forms of
stress concentrations for design purposes. But such stress averaging is possible only in
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

conjunction with ductile materials. As pointed out previously, the brittle nature of high-
strength non-metallic fiber material such as aramid or carbon prevents the utilization of a
large fraction of this strength.
A separate set of problems is posed by the smooth surfaces of aramid fibers and the
resulting low bond strength. This complicates both the stressing operation and the means
of permanent anchorage. In standard prestressed concrete applications, these problems
have been solved by a variety of commercial systems. But it took years to perfect those
systems. It can be assumed that concentrated efforts will likewise result in practical
schemes for high-performance fiber applications.

ACKNOWLEDGMENTS
The authors wish to express their gratitude towards Hexcei-Schwebel for supplying the
fabrics used in this study. Waste glass was kindly provided by Strategic Materials. The
reported progress would not have been possible without the active engagements of Dr. S.
Shimanovich and Dr. S. Kozlova.

REFERENCES
Balaguru P.N., Shah S.P. (I 992). Fiber Reinforced Cement Composites, McGraw-Hill.
Bentur A., Peled A. and Yankelevsky D. ( 1997). "Enhanced Bonding of Low Modulus
Polymer Fibers-Cement Matrix by Means of Crimped Geometry", Cement and Concrete
Research 2 7, 1099-1111.
Berkeley Lab, Operated by the University of California for the U.S. Department of
Energy, http://www.lbl.gov/MicroWorlds/Kevlar/KevlarClue3.html.
Broadway, A. (2002). Hexcel Schwebel. Personal Communication.

Copyright American Concrete Institute

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
28 Vilkner and Meyer
Curbach M. and Zastrau B. ( 1999). ''Textilbewehrter Beton - Aspekte a us Theorie und
Praxis", in Baustatik Baupraxis, Meskouris K., Balkema A.A., Rotterdam, xl-xiO (in
German).
Daniel J.I. and Shah S.P ., eds. (1990). Thin-Section Fiber Reinforced Concrete and
Ferrocement, ACI SP-124.
Gardiner T. and Currie B. (1983). "Flexural Behavior of Composite Cement Sheets Using
Woven Polypropylene Mesh Fabric", Int. Journal of Cement Composites and Lightweight
Concrete 5, 193-197.
Kasperkiewicz, J. and Reinhardt H.W. (1992). "Aramid Fabric as a Reinforcement for
Concrete", in Fiber-Reinforced Plastic Reinforcement for Concrete Structures, A. Nanni
A. and C.W. Dolan, eds., ACI SP-138, 149-162.
Kri.iger M. (2004). "Vorgespannte Dunne Platten aus Textilbeton", Ph.D. Thesis,
University of Stuttgart (in German).
MatWeb, The Online Materials Information Resource, http://www.matweb.com
Meyer C. and Vilkner G. (2003). "Glass Concrete Thin Sheets Prestressed with Aramid
Fiber Mesh", in High Performance Fiber Reinforced Cement Composites 4, A.E.
Naaman and H.W. Reinhardt, eds., E&FN Spon, London.
Mindess S., Bathia N. and Yan C. (1987). "The Fracture Toughness of Concrete under
Impact Loading", Cement and Concrete Research 17, 231-241.
Naaman, A.E. (2000). Ferrocement & Laminated Cementitious Composites, Techno
Press 3000, Michigan.
Peled A., Bentur A. and Yankelevsky D. (1998). "Effects of Woven Fabric Geometry on
the Bonding Performance of Cemehtitious Composites", Advanced Cement Based
Materials 7, 20-27.
Peled A., Shah S.P. and Banthia N., eds (2000). High Performance Fiber Reinforced
Concrete Thin Sheet Products, ACI SP-190.
Ramakrishnan, V., Meyer, C., Naaman, A.E., Zhao, G. and Fang, L. (1995). "Cyclic
Behavior, Fatigue Strength, Endurance Limit and Models for Fatigue Behavior of FRC",
in High Performance Fiber Reinforced Cement Composites 2, A.E. Naaman and H.W.
Reinhardt, eds., E&FN Spon, London.
Reinhardt, H.W. (2002). Concrete Material Science to Application, N. Banthia et al, eds.,
ACI SP-205.
Sabir, B. B., Wild, S., and Bai, J. (2001 ). "Metakaolin and Calcined Clays as Pozzolans
for Concrete: A Review", Cement and Concrete Research 23, 441-454.
Vilkner G. (2003). "Glass Concrete Thin Sheets Reinforced with Prestressed Aramid
Fabrics", Ph.D. Thesis, Columbia University, New York.

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 29

Table 1. Mechanical Properties of Polyamid Fibers (MatWeb)

Nylon 6,6 Nom ex Kevlar 29 Kevlar49 Twaron

Tensile Modulus
3.3 NIA 70 I 12 45
(Gpa)
Tensile Strength
80 300-600 3600 3000 2030
(Mpa)
Ultimate Strain
80 20-30 3.7 2.4 4.5
(%)
Density
1.14 1.38 1.44 1.44 1.45
(J!/cm 3)

Table 2. Mechanical Properties of Roving Systems

Ultimate Tensile Strength Elastic Modulus
Test#
(GPa) (GPa)
8 rovings in 1 row 2.75-3.1 103
34 rovings in 2 rows 1.95 110

Table 3. Glass Concrete Mix Design

Base Materials Weight Ratios

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Glass Agj!Tegate 1.72

Type III Cement 0.8
Metakaolin 0.2
Water 0.38
Liquid Admixture 0.0125

Amid
(trans- and cis- configuration)

:-Jylon 6.6
(a non-aromatic polyamid)

Para-Aramid Fiber (i.e. Kevlar)

Meta-Aramid Fiber (i.e. Nomex)

Figure I. Amid Configurations and Monomers that form Polyamids

Copyright American Concrete Institute

30 Vilkner and Meyer

375 Z700 3.5 - - -~- - - - - 1 - - -:- - - r - -,- -- -,- - '
350 Z450 3
300 Z11JO
l2.5 - - -1- - - - - l ' -- T -

]250
ol
750; .
!:!.. 2
' I

•
~200
~150
! 1.5
.!
'iii
100 c:
~ 0.5 ----,--T--r--,---,--~

50 I
' '
0 '
00 0
0.08 0 2 3 4
STRAIN Strain(%)

a) Stress-Strain Behavior of Various b) Stress-Strain Behavior of

Fiber Materials (Naaman 2000) 8 Aramid Rovings
Figure 2. Typical Load-Deformation Curves

Figure 3. Epoxy Anchorage for Aramid Textile Fabric

4D -------------------- 4.0 - - - -,- - - 1 - - - ,- - - 1 - - - - - - -,

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

-3.5 3.5
z z~
~3.0 3.0 - _I
I

-g 2.5
0
..J 2.0
..
"CJ
0
..J
2.5
2.0
Cl
:; 1.5 .! 1.5
c 'iii
~ 1.0 c 1.0
0.5 ~
0.5
0.0 +--~-----.----r---"'"-,
0.0
0 5 10 15 20 25 0 1 2 3
Experiment Time (min) Strain(%)

a) Load vs. Time b) Load vs. Strain

Figure 4. Response of Aramid Textile Fabric

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 31

12
(83)

~
~
!. 8
=(55)
i.
~
~
~ 4
.3 (28)

0~----~------~----~------~----~
0.00 0.10 0.20 0.30 0.40 0.50
Strain(%)

Figure 5. Glass Concrete Stress-Strain Curve

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

32 Vilkner and Meyer

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

SP-224-03

Textile Reinforced Concrete:

Investigations at Different Levels

by J. Hegger, A. Sherif, 0. Bruckermann and M. Konrad

~ Even though the knowledge about the load bearing behavior of Textile
Reinforced Concrete (TRC) is still limited, there are already applications ofTRC such as
cladding panels and integrated framework systems. Up to the present, the design and
dimensioning ofTRC members is mainly based on extensive test series targeted to the
particular application. Certainly, this approach is very goal-oriented. However, because
design rules are not supported by mechanical models, high safety factors are
incorporated. Within the scope of the collaborative research center "TRC: foundations for
the development of a new technology" (SFB 532) at the Technical University of Aachen,
Germany, the missing consistent description ofthe load bearing behavior ofTRC is being
developed. Thereby, experiments and numerical simulations at different levels, i.e., micro-,
meso- and macro-levels, are performed.ln this paper, the main results of the research
program are presented.

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Keywords: analysis; finite elements; material model; multi-level; textile

reinforced concrete

33
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=University of Texas Revised Sub Account/5620001114, User=erur, ert
No reproduction or networking permitted without license from IHS Not for Resale, 01/26/2015 02:27:05 MST
Daneshlink.com
daneshlink.com

34 Hegger et al.
AUTHOR'S BIOGRAPHIES

Josef Hegger is Professor at the Civil Engineering Department, RWTH Aachen,

Germany since 1993. He obtained his Ph.D. 1985 at TU Braunschweig. From 1985 until
1993 he was employed by the German civil contractor Phillip Holzmann. His main
research interests include shear, high-strength concrete, and textile reinforced concrete.
He is conveyer of the fib task group TG 4.2 Ultimate Limit State Models.

ACI member Alaa Sherif is Assistant Professor in the Civil Engineering Department,
Helwan University, Cairo, Egypt. He obtained his Ph.D. from the University of Calgary,
Canada in 1996. He is an Associate Member of ACI Committee 352 Joints and
Connections in Monolithic Connections. His main research interests include the behavior
and serviceability of reinforced concrete structures.

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
Oliver Bruckermann is research assistant at the Institute for Structural Concrete at the
RWTH Aachen, Germany. He obtained his Diploma Degree in the field of structural
engineering from the RWTH Aachen in 1999.

Martin Konrad is research assistant at the Chair of Structural Statics and Dynamics at
the RWTH Aachen, Germany. He obtained his Diploma Degree in the field of structural
statics from the RWTH Aachen in 2002.

RESEARCH SIGNIFICANCE

The reinforcement of concrete with technical textiles extends its application to

completely new fields. Because of the corrosion resistance of the textile materials, thick
concrete covers as known in ordinary reinforced concrete are no longer needed. Thus,
slender structural members with a wall thickness as low as ten millimeters are possible.
In addition, fine grain concrete matrices guarantee an even and sharp-edged high quality
surface, so that TRC is predestinated for architectural applications. Although the
knowledge about the load bearing behavior of TRC is still limited, applications such as
cladding panels [ 1] and an integrated framework system [2] have already been
implemented.

The material models currently being developed and presented in this paper will lead
to a better understanding of the failure mechanisms of TRC-structures. Based on these
models, design rules and safety concepts can be set up in the future. Not until then TRC
will become a versatile and cost-effective building material.

INTRODUCTION

Current investigations show that the failure process of TRC, which is observed at the
macro level, cannot accurately be predicted based on the simple models known from
ordinary steel reinforced concrete. This is primarily due to the inhomogeneous internal
structure of the fiber strands (rovings) consisting of hundreds of separate filaments
(Fig. I). Furthermore, non-uniform bond conditions in the longitudinal direction and

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 35

within the cross-section of the rovings point out that developing a material model,
capable of accurately describing the load bearing behavior of TRC is a non-trivial task.

Therefore, within the SFB 532 the investigations are being conducted at different
levels. As shown in Table 1, these are the micro-, meso- and macro-levels. At each level,
the material components have to be defined and modeled by an appropriate Finite
Element model. Corresponding experiments provide the necessary data for calibrating the
numerical models and obtaining the required mechanical properties of the components.
The models at each level may conceptually either be coupled to an adaptive multi-level
computation. Or as an alternative, the "smeared" material parameters used at one level
are predefined at the previous smaller level. In the following, the main results for each
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

level are described.

MICRO-LEVEL

The aim of the micro-level investigations is to completely cover the effects

determining the behavior of single filaments as well as their mutual interaction and their
interaction with the concrete matrix. In addition to the material properties, such as tensile
strength and bond-slip relationship, the random distribution of several parameters has to
be considered. Among these are the following [3]:
• Micro defects in the filament which lead to a statistical size-effect
• Cross-section of the filament
• Geometrical position of filaments within the roving cross-section and the change
in the longitudinal direction (waviness)
• Bonding conditions, i.e., to what degree does matrix contact the filament's
perimeter and how does this vary in the longitudinal direction

The interaction between filament and matrix has been studied using pullout tests of
single filaments (Fig. 2(a)) and the finite element model shown in Fig. 2(b). The resulting
bond-slip relation (Fig. 3) is in agreement with the one obtained analytically by
Banholzer [4, 5] based on the shear lag theory. The bond-slip curve can be divided into
three parts. The first part is determined by elastic adhesional bond. Here, the bond layer is
capable of taking up stresses of up to 4.8 N/mm 2 at a very small slip of 0.01 mm. If the
value of the slip exceeds 0.0 I mm, the adhesional bond is lost and only friction bond
remains which is about four times smaller than the adhesional bond. In the second part
(slip from 0.01 to 0.16 mm) the friction bond degrades due to "smoothing" of the
interface between filament and matrix. In the third part the friction bond reaches a
minimum and remains constant. This behavior is very different from steel reinforcement,
because the maximum bond-stress can only be activated at a certain value of slip, i.e.,
there is no ductility. Therefore, the maximum force picked up by the filament, cannot be
calculated just by the maximum bond stress times the available anchorage length.

Copyright American Concrete Institute

36 Hegger et al.
MESO-LEVEL

The meso-level investigations aim at describing the behavior of the rovings within
the composite material. Pullout tests, of which several types exist, are the main tools to
study the characteristics of the bond between roving and matrix (Fig. 4). In contrast to
steel reinforcement, the load-displacement curve does not feature a plastic plateau.
Rupture of filaments or debonding over the entire embedment length lead to a rather
quick decrease of the pullout force after the peak-load.
In order to simulate pullout tests the so-called bond layer model shown in Fig. 5 has
been developed. This model considers a section of the roving idealized as layers and
takes into account the deterioration of the bond with increasing distance between filament
and matrix. The bond quality for each layer is defined by a function of the distance from
the matrix. Any changes of these proportions in the longitudinal direction are neglected.
In the Finite Element simulation each bond layer is represented by a one-dimensional
element and is connected to the matrix by a zero-thickness interface element.

The parameters to be calibrated for the bond layer model are the bond quality
distribution function and the effective tension strength of the filaments. In Fig. 6(a) the
influence of a linear, a quadratic and a cubic bond quality distribution on the load-
displacement curve of a roving pullout test is shown. Figure 6(b) indicates the calculated
actual fraction of unbroken filaments. In the pullout experiment, this information was
obtained by optical recording of the light transmission through the unbroken filaments
[6). Both, the linear and the quadratic approaches cannot accurately reproduce the smooth
peak region of the measured curves. The cubic distribution leads to a more ductile curve
and resembles the experimental data.

While the bond quality distribution function determines the post peak gradient, the
maximum pullout force depends mainly on the tensile strength of the filaments.
Preliminary calculations showed that the initial slope of the pullout curve is always
overestimated by the model. Only a decrease in the maximum bond performance can
reduce the initial stiffness, however this leads to a simultaneous decrease of the portion of
filament fracture, because more filaments are pulled out completely. Therefore, the
variation range of this parameter is limited, and the initial stiffness in the experiments
cannot be reproduced by solely reducing the bond quality. As a consequence, this
reduction can be explained by the existence of an internal free length between the
macroscopic boundary of the matrix and the first contact of the filaments with the matrix
inside the specimen, i.e., the start of the micro-bonding between filament and matrix,
which is illustrated in Fig. 7. Using a cubic bond quality distribution function, a tensile
strength of 1350 MPa and a maximum internal free length of 6 mm in the center of the
roving, the simulation results in the pullout curve shown in Fig. 8.

MACRO-LEVEL

The tension tests on composite specimens reinforced with rovings and its simulation
play an important role in the theoretical models. Based on the results of the pullout test
described above (meso-level), the target is to explain the macro-response, i.e., the load
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 37

displacement curve of TRC under tension load. Pure rovings are used instead of textiles
in order to avoid any influences resulting from the textile production process, the textile
joints, and from the lateral reinforcement. Fig. 9 illustrates the geometry of the tensile test
specimens and Fig. 10 shows the load-strain curve for the measurement length of a
specimen reinforced with 12 rovings (2400 tex each). At the beginning, the load strain
curve follows the usual linear-elastic increase. At a force of 3.1 kN the first crack occurs.
This very early cracking can be explained by an initial bending moment in the specimen
due to a small load eccentricity, which practically cannot be avoided. At a force of 4.8 kN
the "real" cracking process starts, i.e., now the lower limit of the tensile strength of the
concrete matrix within the specimen is reached ( 4.8 N/mm 2). Until a strain of 6 %o there
is a continuous increase in the force and the number of cracks. The maximum load
reaches 7.2 kN after which macroscopic failure can be observed. The specimen features a
very satisfactory crack distribution (Fig. 11) with an average crack distance of about
14 mm and crack width of 0.09 mm. This is calculated from the elongation of the
measurement length neglecting any elastic deformation of the non-cracked concrete.

The interpretation of the experimental data raises the following question: Why is
there a substantial increase in the stress while cracking occurs? Normally, one would
expect, similar to steel reinforced concrete, that cracking happens at a stress plateau,
which is only determined by the tensile strength of the matrix.

In order to investigate this phenomenon a Finite Element analysis was conducted

with the simple model shown in Fig. 12 (a). The model consists of linear elastic truss
elements representing the uncracked concrete and of short crack elements (/ = 1 mm).
The number of crack elements is chosen according to the number of cracks occurring in
the experiment (35 cracks). For the crack elements a material law is used which can be
split into two parts (Fig. 13):
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

• Concrete material law with mesh adjusted softening modulus (7]. The tension
strength varies randomly (average = 5.0 N/mm 2 , standard deviation = 0.25
N/mm 2 ).
• Material law for the reinforcement, whereby the stiffness and the maximum load
( 12 x 430 N = 5160 N) are taken from the linearly approximated pullout curve.
The result obtained is displayed in Fig. 14 (res_wo_core). Because the maximum load
that can be picked up by the reinforcement is higher than the crack load, all 35 cracks
open up. Finally, one of the cracks localizes and the reinforcement fails. The strain at
failure of 6 %o matches the experiment. However, the model is not capable of predicting
the linear load increase with increasing number of cracks.

In order to model this behavior, it is necessary to introduce a "core" of filaments,

which are activated gradually with increasing crack number and strain, respectively (cf.
[8]). This delayed activation is due to the following:
• Filaments in the center of the roving cross-section may have a waviness. Before
they pick up load, they have to be stretched out.

Copyright American Concrete Institute

38 Hegger et al.
•At the crack edge the core filaments do not have direct contact to the matrix.
Therefore, their anchorage length is much longer than of those (sleeve-)
filaments taking up the pullout force.
The bond force of the core is transmitted to the matrix not so much through friction bond
between adjacent filaments as through matrix which has penetrated the roving at several
points. These considerations lead to the modified Finite Element model shown in
Fig. 12 (b). Of course, the introduction of a core, which is activated through bond
elements, actually introduces several new unknown parameters, such as cross-section,
stiffness and tension strength of the core. In addition, the friction law applied has to be
defined. A curve-fitting ofthe experimental data leads to a portion of35% core filaments
related to the whole roving cross-section and a tension strength of 600 N/mm 2 • The
increase of bond flow is 65 N/mm per 1 mm of slip. The load strain curve now matches
the experimental curve much better (see res_w_core in Fig. 14).

It must be kept in mind that the so far obtained material laws may not be the only
solution leading to these results. Further investigations on tension-tests with different
lengths and different degrees of reinforcement must provide additional data for
calibrating the model.
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

SUMMARY

The explanation of the load bearing behavior of TRC is a task that only can be
handled if different resolution levels in the experiments as well as in the simulation are
considered. In this research the behavior of TRC is investigated at three different levels:
micro-, meso- and macro-level. It is shown how results obtained at one level can be used
at the next higher level in order to reduce the number of unknown parameters. A finite
element model is presented, and first results show its capability to simulate the
experimental results. However, systematic experimental research has to provide further
information in order to verifY and optimize the models at each level. In addition,
simulation techniques have to be developed at the macro-level taking into account the
influences of the textile production process and the specific geometry of textiles. The
demonstrated simulation of roving tension tests is only an intermediate step towards this
final goal.

ACKNOWLEDGEMENTS

The authors thank the German research foundation (DFG) in context of the Collaborative
Research Center 532 for their financial support.

REFERENCES

I. Hegger, J., Schneider, H. et al., "Exterior Cladding Panels as an Application of Textile

Reinforced Concrete", this journal.

2. Brameshuber, W. et al: "Textile Reinforced Concrete (TRC) for Integrated

Formworks", Proceedings ofTechtextil-Conference, CD-ROM, Frankfurt, 2003.

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 39

3. Chudoba, R., Vorechovsky, M., Konrad, M., "Stochastic modelling of multifilament
yams focused on the delayed filament activation and size effect", in preparation.

4. Banholzer, 8.; Brameshuber, W., "Eine Methode zur Beschreibung des Verbundes
zwischen Faser und zementgebundener M~trix", Beton- und Stahlbetonbau, No. 96, pp.
663-669,2001 (in german).

5. Cox, H.-L., "The elasticity and strength of paper and other fibrous materials", British
Journal of Applied Science, Vol3, pp.72-79, 1952.

6. Brameshuber, W. ; Banholzer, B. ; Pierkes, R., "Investigations on Bond Characteristics

of Textile Reinforced Concrete Using the Confocal Laser Scanning Microscopy",
Proceedings of the 8th Euroseminar on Microscopy Applied to Buildings Materials,
Athen, September 4-7, pp. 533-540, 2001.

7. Jirazek, M., Bazant, Z., "Inelastic Analysis of Structures", John Wiley & Sons, 2002.

8. Ohno, Sadatoshi, Hannant, "Modelling the Stress-Strain Response of Continuous Fiber

Reinforced Cement Composites", ACI Materials Journal, pp. 306-312, 1994.

CONVERSION FACTORS

I in. = 25.4 mm
I ft = 0.3048 m
I kip = 4.448 kN
I lex =lg/km

Table 1 --Considered components at different resollllionlevels

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Resolution level Components

Micro-level • filament
• matrix
• bond filament-matrix
• bond filament-filament

\tfeso-levcl • subroving (group of filaments)

• matrix
• bond subroving-matrix

Macro-level • roving I textile

• smeared concrete
• smeared bond roving-matrix

Copyright American Concrete Institute

40 Hegger et al.

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
Fig. 1- X-section of a roving embedded in fine concrete

Cone tete elements. EA"' co

-
<'[NJ
A (mm) ~--"'1---...-....,;;........--~

X ~
.,.._,..,...,......._,....F,;\

\ filament
y'
splice< t\,01 rnm Filament elements

(aj Pullout test (h) Finite element model

Fig. 2 - Filament pullout test and Finite Element model
~
6
"E '

-
'
E 5 '
--4,8 '
2:. '
'
tl) 4 '
.. LIII
,-
tl)
~
G)

3 [!1/1
II /I III
t l)

c I '
'
0 2 '
.Q '
'
0,5 '
'
'
'
0
0,1 0,2 0,3 0, 4
0 I_

0,01
I
0,16 slip [mm]
Fig. 3 - Bondstress-slip diagram for a single filament embedded in fine concrete

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 41

gooo~--------------------~
"'·.;oo
epoxy resin

F[N}
~~~~~------~--~
u{mm}
---+

0.0 0,5 1.0 l,S 2,()

~u(mm}

Fig. 4 - Roving pullout test and typical load displacement curve

decreasing
. bond quality

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
o: o.• o.e •.o •
distance flom yam perimeter

(a! ideulized roving x-sectimt (b) bond quali(l' distributi(m.funclions

Fig. 5 - Bond layer model

roo _,oo
t.
j 8D

i
"& 20
40

0.1 0.2 0.3 0.4 0.5

J 0~----T-----T-----~----~--~
0 0,1 0.2 0.3 0.4 0.5
displacemont [mmJ dltplac......,ll"""l
(a) on the load-displacement curve (b) on the fraction l?(unbroken filaments
Fig. 6 - Influence of the bond quality distribution function

Copyright American Concrete Institute

42 Hegger et al.

Fig. 7 - Internal free length between matrix boundary and begin of microbonding

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
..... 600 .
~
e 500
.e
400'

300

200.

100

0
0 0,2 0,4 o,a o.a
displacement (niml
Fig.8 - Roving pullout test, experiment vs. simulation

200 500 200

rovings
[mm]
!width= 100 mmj
Fig. 9 - Geometry of tensile tests

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 43

-
-
.......
6,0
..A...N"'\,. v-w
.---- ~ -.......
5,0

4,0 r
3,0 7
2,0

1,0

0,0
0,0 1,0 2,0 3,0 4,0 5,0 6,0 7,0
strain l'J'e)
Fig. 10- Load-strain curve of a tension specimen reinforced with 12 rovings

Fig. II - Crack distribution of a tension specimen reinforced with 12 rovings

·-
nonlinear [pl~stiC) bond elements
F
~ ijilj f'·*:;j:11[6'l~i;[~i;li!': [l!f:::j;;jct~i·I:;J;1{-+
~ ! /
~;.-'

cae elemen!S
a! FE- modd with huilt·in cracks fhJ FE-mndel Wilh core added

Fig. 12 - FE - model at macro-level

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

44 Hegger et al.
~500
u.
B400+-~~--r-~----r-------~----~

5
u.300~~~~~------~--------~----~

2·G1 0,0 0,5 1,0 1,5 2,0

L·ft displacement u [mm]
a) Concrete tension law (b) Approximated pullout curve
Fig. 13 - Materia1laws of crack elements

z~8r----------------------------------------,
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

~
;7+---------~----------------------~~~~~~

~
-6r---------~~~~~~----~~~~

3 -res_wo_oore
-res_w_core
2
[ . -experiment

0+---------------------------------~----~
0 2 4 6
strain [%o]
Fig. 14 - Tensile test, experiment vs. simulation

Copyright American Concrete Institute

SP-224-04

Textile Reinforced Concrete (TRC) for

Integrated Formworks

by W. Brameshuber, M. Koster, J. Hegger, S. Voss, T. Gries,

M. Barle, H.-W. Reinhardt, and M. Kriiger

S)'D~ This paper presents use of textile reinforced concrete (TRC) for producing
integrated formwork element for use in construction. The TRC integrated formwork
elements are significantly lighter compared to the normal precast elements owing to their

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
relatively smaller wall thickness, typically around I 0 mm. The cross-section ofthe TRC
integrated formwork element can be chosen as dictated by the specific application and the
composite can be designed to have a high load-bearing capacity. The fresh concrete is
protected against moisture loss by the integrated formwork elements that remain in place
in actual construction. Hence, neither demolding of the TRC integrated formwork nor
curing of the poured concrete is required. The TRC integrated formwork elements also
possess the advantage of having a surface appearance of high quality. This contribution
presents a compilation of the results from the testing performed on the TRC integrated
fonnwork elements.

Keywords: concrete; formwork; textile reinforced concrete

45
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=University of Texas Revised Sub Account/5620001114, User=erur, ert
No reproduction or networking permitted without license from IHS Not for Resale, 01/26/2015 02:27:05 MST
Daneshlink.com
daneshlink.com

46 Brameshuber et al.
Wolfgang Brameshuber is Chair of the Building Materials Science/Structural Materials at
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

RWTH, Aachen University. He was employed at BUNG Engineering Office, Heidel-

berg, Germany, Central Laboratory at Bilfinger Berger, Mannheim, Germany. He ob-
tained his doctoral degree from University of Karlsruhe, Germany in 1988.

Matthias Koster is Research Associate at the Institute of Building Materials Research at

RWTH Aachen University, Germany. He received his Diploma Degree in physics from
the University of Bonn in 1996 and a Diploma Degree in the field of civil engineering
from the University of Applied Sciences, Trier in 2000.

Josef Hegger is Professor at the Civil Engineering Department, RWTH Aachen Univer-
sity, Germany. He obtained his Ph.D. in 1985 from TU Braunschweig. He was employed
by the Gennan civil contractor Phillip Holzmann, 1985-1993. His main research interests
include shear, high-strength concrete, and textile reinforced concrete. Conveyer of the fib
task group TG 4.2 Ultimate Limit State Models.

Stefan Voss is Research Associate at the Structural Concrete Department at RWTH

Aachen University, Germany. He obtained his Diploma Degree in the field of structural
engineering from Aachen University in 2002.

Thomas Gries is Professor at the Institute for Textile Technology, Aachen. He obtained
his doctoral degree in 1995. He was employed with Lurgi Zimmer AG, Frankfurt a.M.,
1995-2000 and has served as the Head of the Department of Technologies for Fibres and
Textiles. He is Chair and Head of the Institute for Textile Technology, Aachen since
April 2001.

Marijan Barle is Research Associate at the Institute of Textile Technology, Aachen Uni-
versity. He obtained his diploma degree in the field of structural engineering from RWTH
Aachen University in 2000.

Hans-Wolf Reinhardt is Professor and Chair of Construction Materials at the Stuttgart

University and Managing Director of Otto-Graf-Institute, Stuttgart, Germany since 1990.
He obtained his doctoral degree from Stuttgart University, Germany in 1968. He was
post-doctoral researcher at liT, Chicago, USA, 1969-1970. He was teaching and a re-
search engineer at Otto-Graf-Institute, Stuttgart, Germany, 1970-1975. He was Professor
and Head of the Stevin Laboratory for concrete structures at Delft University of Technol-
ogy, Netherlands and Professor for Building Materials and Building Physics at Darmstadt
University of Technology, Germany, 1975-1986.

Markus Kruger is currently working on his PhD project "Prestressed Textile Reinforced
Concrete" at the Institute of Construction Materials, University of Stuttgart. He gradu-
ated from the University of Dortmund in civil engineering in March 1998. He has been
engaged in teaching and educating students in concrete technology and in the develop-
ment and application of non-destructive test methods in civil engineering.

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 47

INTRODUCTION
It is generally possible to cast elements for moulds using textile reinforced concrete as
demonstrated elsewhere (I ,2). This new technology allows the use of very light elements
which remain in the structure after casting the concrete on site. Several benefits result due
to the use of the integrated formwork for concrete structures. Integrated formworks are
very light due to the small thickness of the textile reinforced concrete elements. Due to
their lightweight feature, it is possible to easily move and fix the integrated formwork in
the field. The visible surfaces ofthe elements possess high quality in terms of roughness,
color and absence of voids compared to normal concrete. The matrix of the textile rein-
forced concrete is very dense leading to high durability. Due to the self-leveling proper-
ties of the matrix, the surface has less visible pores and is very smooth compared to nor-
mal vibrated concrete. From economic considerations, the aspects of formwork demold-
ing and curing of poured concrete deserve special attention. Since the integrated form-
work becomes integral part of the structure, the need for its demolding and the need for
the curing of the poured concrete are eliminated.

As a first step in this investigation, textile reinforced concrete integrated formwork ele-
ments were developed to understand the load-deformation behavior of the concrete ma-
trix and reinforcing textiles chosen as material variables. Some results from the investi-
gation are given in References 1 and 2. A design model has been developed to calculate
the load-carrying capacity of the formwork having U-shaped cross-section. In Fig. I, the
predictions from the design model are compared to experimental values. The calculated
values agree well with the experimental data except for the combination of concrete
83/textile WI. For this combination, no bending failure was observed. The failure mode
in this case was dictated by the break down of the connection between the flanges and the
plate of the U-shaped structure. Therefore, the calculation based on bending failure re-
sults in higher values compared to the experimental ones.

Based on these results, a new geometry of textile reinforced concrete integrated form-
work has been developed. The new integrated formwork and the corresponding testing
results are described in this contribution.

MATERIALS
The concrete mixture composition used in this investigation is described elsewhere in
Reference 2. The cementitious matrix used for producing the integrated formwork has a
high llowability due to presence oflarge volume and dense network of the reinforcement.
The cementitious matrix used in this investigation has a relatively high Youngs modulus
of up to 35000 N/mm 2, and the modulus ofrupture of about 5 N/mm 2•
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Examples of the textiles used as reinforcement in the integrated formwork are given in
Fig. 2. In this case, the so-called friction spun hybrid yams have been woven to obtain a
2-dimensional textile structures. The yarns consist of alkali-resistant glass for the core
material and polypropylene for the sheet material. In view of the composite mechanical
behavior, most influential are the thickness and the number of yams in the core of there-
inforcement. Melting the polypropylene after producing the textiles gives a good inner
bond between the yarns and a relatively high tensile strength of the textile.

Copyright American Concrete Institute

48 Brameshuber et al.

For producing the integrated formworks, 3-dimensional textile structures of reinforce-

ment may be used. Therefore, the 2-dimensional textiles are connected with spacers, for
example, spacers consisting of aramid. The 3-dimensional textiles can be produced in a
single step using special manufacturing techniques.

Tensile Tests
The mechanical behavior of the textile-concrete composite was verified using the uniax-
ial tensile test. Test set-up and geometry of the specimen are shown in Fig. 3. Examples
of the load-strain-curves obtained in the testing are shown in Fig. 4. In this figure, the ef-
fectiveness of textile with friction spun hybrid yarns (FSHY) over normal textile is evi-
dent. It can also be observed that increasing the glass content in this type of yam (FSHY)
leads to a higher composite stiffness and a slightly higher load to failure.

DESIGN OF CROSS-SECTION
The first generation of U-shaped integrated formworks had a span width of 1.5 m. For
practicable application, it is more suitable to have longer elements with supports in be-
tween. The new elements therefore should be able to carry positive as well as negative
moments. Therefore, a cross-section has been developed as shown in the figure appearing
in Table 1. Using the design model developed, a pre-calculation had been conducted to
estimate the required element cross-section. The results for dependence of the span width
as well as the content of textile reinforcement on the element cross-section are shown in
Table 1. Note that one layer of reinforcement represents about 245 mm 2 per meter width
of the reinforcement.

Based on practical considerations, the integrated formwork with a triple span width of
3x 1.5m has been chosen for testing. By making some modifications in the textile, it was
possible to reduce the inner lever-arm of the element from 60 mm to 50 mm when using
two layers of textiles. It should be noted that use of three layers of textiles is not suitable
in practice due to the danger of entrapping large voids in the concrete during casting. The
chosen cross-section has the advantage of carrying high bending moments in two direc-
tions. Additionally, there is a good physical bond to the site concrete due to the presence
of flanges having a slight negative slope. The required rheology performance of the con-
crete is very demanding due to the complex geometry of the element. Fig. 5 shows a
photograph of the new integrated formwork. It can be observed that the textile is ar-
ranged very precisely in the middle of the walls and no defects occur on the concrete sur-
face.

TEST RESULTS
Fig. 6 shows an example flexure load-deflection curve obtained from the testing. This
data shows that the maximum load and the stiffness after cracking for the tested compos-
ite were somewhat less but comparable to the old formwork element (2) with bigger
cross-section. It should be noted that the old element was 120 mm high compared to the
new element with a height of60 mm only.

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 49

The elongation of the bottom side of the element between the load-points was measured
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

using an appropriate device. This measurement provides the opportunity to calculate the
strain on the tension side of the specimen and the strain in the textile as a mean value af-
ter making some geometrical corrections. Further, a mean value for the crack-width in
dependence to the load can also be calculated. This type of information is very important
for understanding the performance of new composite materials such as textile reinforced
concrete. Fig. 7 shows the load-strain curve corresponding to Fig. 6.

Textile reinforced concrete works suitably if the crack widths and the crack spacing are
small associated with the large number of cracks. In the above example, high strains are
observed with the formation of cracks having spacing Jess than I 0 mm. This behavior
demonstrates the good rotation capacity of the TRC composite.

SUMMARY
This paper presented use of textile reinforced concrete (TRC) for producing integrated
formwork element for use in construction. The TRC integrated formwork elements are
significantly lighter compared to the normal precast elements owing to their relatively
smaller wall thickness. The cross-section of the TRC integrated formwork element can
be chosen as dictated by the specific application and the composites can be designed to
have a high load-bearing capacity. The fresh concrete is protected against moisture Joss
by the integrated formwork elements that remain in place in actual construction. Hence,
neither demolding of the TRC integrated form work nor curing of the poured concrete is
required. The TRC integrated formwork elements also possess the advantage of having a
surface appearance of high quality. This contribution presents a compilation of the re-
sults from the testing performed on the TRC integrated forrnwork elements.
The integrated forrnwork has been developed as an applicable construction element. Nev-
ertheless, the research work is in progress and the main topics of investigation include
understanding the interaction between the integrated forrnwork and the site concrete.
Need also exists to understand the improvement in fire resistance performance obtained,
if any, as a result of the use of integrated formwork in construction applications.

ACKNOWLEDGEMENTS
This work (A iF-No. 47 ZN/DBV 229) is supported by the German Association of Con-
crete and Building Technology (Deutscher Beton- und Bautechnikverein E. V.) and the
Union of Industrial Research Association (Arbeitsgemeinschaft industrieller Forschungs-
vereinigungen) with financial resources of the Federal Ministry for Economy and
Technology.

LITERATURE
(I) Hegger, J. ; Sasse, H.R. ; Wulfhorst, B. ; Doinghaus, P. ; Ro131er, G. ; et al: U-
Shaped Supports as Formwork Elements Integrated in the Construction Member.
Frankfurt : Messe, 1999. - In: TechTextil Symposium lnnovatives Bauen 5.1
Textilbewehrter Beton - Material und Produkte, Frankfurt, 13. April 1999,
Vortrag 517, 8 Seiten

Copyright American Concrete Institute

50 Brameshuber et al.
(2) Brameshuber, W.; Brockmann, J. ; RoBier, G.; Hegger, J.; et al: Textile Rein-
forced Concrete for Formwork Elements. Frankfurt : Messe, 2001. - In: II.
lnternationales Techtextil-Symposium fiir technische Textilien, Vliesstoffe und
textilarmierte Werkstoffe, Frankfurt, 23.- 24.04.2001, Vortrag 33570917

(3) Brameshuber, W.; Koster, M. ; Hegger, J. ; Voss, S. ; Gries, T. ; et al: Textile

Reinforced Concrete (TRC) for Integrated Formworks. Frankfurt: Messe, 2003.
-In: 12th International Techtextil Symposium for Technical Textiles, Nonwov-
ens and Textile Reinforced Materials, Frankfurt, 7-10 April, 2003, Paper 4.23, 6
pages

(4) Hegger, J. ; Bruckermann, 0. ; Voss, S. ; Brameshuber, W. ; Brockmann, T.:

Decentralized Wastewater Treatment Plants Made of Textile Reinforced Con-
crete. Frankfurt: Messe, 2003.- In: 12th International Techtextil Symposium for
Technical Textiles, Nonwovens and Textile Reinforced Materials, Frankfurt, 7-
10 April, 2003, Paper 4.28, 38 pages

(5) Hegger, J.; Molter, M.; Will, N.: Facades made out of Textile Reinforced Con-
crete. Leipzig Verein der Freunde des Bauingenieur- und
Wirtschaftsingenieurwesens an der Universitiit Leipzig, 2002.- In: Proceedings
of the 6th International Symposium on Utilization of High Strength/High Per-
formance Concrete (Konig, G. ; Dehn, F. ; Faust, T. (Ed.)), Leipzig, June 2002,
Volume 1, S. 125-136

(6) Brameshuber, W.; Brockmann, J.; RoBier, G.: Textile Reinforced Concrete for
Formwork Elements - Investigations of Structural Behaviour. London : Thomas
Telford, 2001. - In: Proceedings of the 5th International Conference on Fibre-
Reinforced Plastics for Reinforced Concrete Structures, Cambridge, 16-18 July
2001, (Burgoyne, C.J. (Ed.)), Vol. 2, S. 1019-1026
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 51

Table I: Design of cross-section

z[J"!:iL~ ~;:;I M,•~•F,

Loading (DIN 4421)
Fresh cot~aetll: 26 kNim"
Wodcial coodltiol\: U kNhn"
90 180 90 -d:i.Fct ~=Z.xFct Scal'futdl foriollds,grollp 1: yT= 1.25
! ; 360 f r (mmj Fc=fct Thlcknau of eoner.t•: 250 mm

-+ o•1s kNim
rmxM, Minimum lever-ann z, [mm)
~~ ~--------~~~~~--------~
Reinfoo::ement
System y: 1.5 1---:-:-----.--::-:-----.---=-:--------i
11ayer 21ayers 31ayers

1.03 55 27 18

J • .A 0.83 84 42 28
I 1~ , \2!\ • 125 I

1 19 12 60 40
150 1SO 1!10

M....ll'i~kl'llttn
600 . , - - - - , - - - - - , - - - - - : - - -

•••
•••
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Fig. I: A comparison of calculated fracture moments versus experimental data for

integrated formworks

Materials

·Glass for core material

- Propylene for sheet material
- Aramld for spacer yarns

•
.~c. .
Fig. 2: Production of 3-dimensional textiles

Copyright American Concrete Institute

52 Brameshuber et al.

A-A
.;;;;;;;, ~~~0

Fig. 3: Test set-up for the uniaxial tensile tests

0+---~--~----~--~--~--~----~~
0 2 3 4 5 6 7 8
Strain (mm/m]

Fig. 4: Load-strain curves from tensile tests

Fig. 5: Cross-section of a cast form work element

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 53

-10
-9 /
-8 /
-7
./
~

/
:z-s / v \
1!.-s
~ -4
.L"¥' I-"' .,...._.,...
4711mm ~
~

~
-3 '---
./
-2
-1 r 1500mm
I--
I--
0
0 5 10 15 20 25 30 35 40 45
Deflection [mml

Fig. 6: Load-deflection curve

-10

z.!.
-9
-8
-7
-6 /
/
//
/
""
'tl -5 -.../'
/
~
.9"' -4 .,...._.,...
470mm
.,..r..J [
-3
/ ! !
-2
•1 /' 1 - 1500mm
~-
:-
0
0 2 3 4 56 7 8
Strain ['L]

Fig. 7: Load-strain curve of bottom flange

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

54 Brameshuber et al.

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

SP-224-05

Exterior Cladding Panels as an Application of

Textile Reinforced Concrete

by J. Hegger, H. Schneider, A. Sherif, M. Molter, and S. Voss

Synopsis: The composite material textile reinforced concrete (TRC) offers a number of
advantages, in particular for the manufacturing of fa~ades. The textile reinforcement and
the possible thin concrete cover, enable the construction of thin-walled structural
components. Filigree cladding panels made of textile reinforced concrete open up new
ways for an entirely new application of the construction material concrete and give
architects and engineers more freedom in the design. In this paper some basic information
about the load bearing behavior of textile reinforced concrete is given and the use of
textile reinforced concrete in a pilot project for the exterior claddings of the extension of
the laboratory hall at the RWTH Aachen University, Germany, is described.

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Keywords: cladding panels; curtain wall; facades; textile reinforcement

55
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=University of Texas Revised Sub Account/5620001114, User=erur, ert
No reproduction or networking permitted without license from IHS Not for Resale, 01/26/2015 02:27:05 MST
Daneshlink.com
daneshlink.com

56 Hegger et al.
AUTHOR'S BIOGRAPHIES

Josef Hegger is Professor at the Civil Engineering Department, RWTH Aachen,

Germany since 1993. He obtained his Ph.D. 1985 from the TU Braunschweig. From 1985
until 1993 he was employed by the German civil contractor Phillip Holzmann. His main
research interests include shear, high-strength concrete, and textile reinforced concrete.
He is conveyer of the fib task group TG 4.2 Ultimate Limit State Models.

Hartwig Schneider is Professor at the Architectural Department for Construction and

Design, RWTH Aachen, Germany since 1999. He studied at the University of Stuttgart in
Germany and the Illinois Institute of Technology in Chicago. Since 1988 he is partner of
an Architectural Consulting Office in Stuttgart and member of the German Association
for Architects since 1993.

ACI member Alaa Sherif is Associate Professor in the Civil Engineering Department,
Helwan University, Cairo, Egypt. He obtained his Ph.D. from the University of Calgary,
Canada in 1996. He is an Associate Member of ACI Committee 352 Joints and
Connections in Monolithic Connections. His main research interests include the behavior
and serviceability of reinforced concrete structures.

Matthias Molter obtained his Diploma Degree in the field of structural engineering from
the Technical University of Darmstadt in 1993. From 1996 to 1998 he worked as a
research assistant at the Structural Department at the University of Karlsruhe, and from
1998 to 2002 at the Structural Concrete Department at the RWTH Aachen, Germany.
Since 2002 he is director of the technical office, Bremer AG, Paderbom, Germany.

Stefan Voss is research assistant at the Structural Concrete Department at the RWTH
Aachen, Germany. He obtained his Diploma Degree in the field of structural engineering
from the RWTH Aachen in 2002.

RESEARCH SIGNIFICANCE

Textile reinforced concrete enables the construction of thin-walled structural

components, because only a thin concrete cover of the textile reinforcement is needed to
ensure the bond between fibers and concrete [I]. Reducing the thicknesses of the
components results in more economic fayade anchors and further load bearing elements.
In a research project, cladding panels made of textile reinforced concrete have been
developed for their application in a pilot project. The design of the panels is based on
experimental investigations on textile reinforced concrete samples with different
reinforcement properties. In this paper some general information are given about the load
bearing behavior of textile reinforced concrete, important test results and a first
application ofTRC are described.

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 57

INTRODUCTION

The development of textile reinforced concrete is based on the fundamentals of glass

fiber reinforced concrete with short fibers. In order to increase the effectiveness of the
fibers, they are aligned in the direction of the tensile stresses, thus, similar to the case of
ordinary reinforced concrete elements. By replacing the ordinary steel reinforcement by
textile reinforcement, filigree-cladding panels with a broad range of design options can
be created. Profile thicknesses, previously known only from steel construction and
composite fiber plastics structures, can be achieved with textile reinforcement as well as
high quality homogenous surfaces. These advantages lead to an entirely new application
potential for concrete as a building material, especially for fac;ade construction. The small
panel thickness of 25 mm achieved by textile reinforced concrete, compared to the 70 to
100 mm required by ordinary reinforced concrete panels, results in a lower dead load and
eliminates the need for complex fa9ade anchors.
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Textile reinforced cladding panels have been used for the extension building of the
Institute of Structural Concrete, RWTH Aachen University, shown in Fig. 1. The existing
single-aisle hall with a span of 12.0 m has been extended by four axes of 5.4 m spacing
each. Curtain wall elements were used except for the lower part (socket) of the building,
where sandwich plates of 35 mm thick facing shells were placed (Fig. 2). Innovative,
textile reinforced concrete components have been developed for this purpose. On the
longitudinal side of the hall, 2685 x 325 x 25 mm curtain wall panels as shown in Fig. 3
have been applied instead of hitherto natural stone, which would have been the typical
choice. The high cost of the natural stone and its manufacture restricts its use to high
quality administrative buildings. Textile reinforced cladding panels are notably less
expensive and are therefore a cost efficient alternative for residential and commercial
structures.

EXPERIMENTAL INVESTIGATIONS

Extensive experimental and theoretical investigations are currently carried out at the
RWTH Aachen to determine the load bearing behavior of textile reinforced concrete.
Therefore, the effect of different fiber materials (e. g. alcali resistant glass (AR Glass),
carbon and aramide), different fiber bundle (roving) and fabric geometries, coatings and
concrete properties are tested. ·

Materials: AR Glass fiber fabrics have been chosen as reinforcement for the fa9ade
panels because of their lower costs compared to fabrics made out of carbon or aramide
fibers. The textile reinforcement fabrics for the test series (Table I) were designed and
manufactured by the Institute of Textile Technology, RWTH Aachen University. They
differ in the roving thickness and the mesh size. The tensile strength C!mm· is determined as
the average value of I 0 tensile tests on 125 mm long parts of rovings taken from the
fabric. The loading has been applied with a deformation rate of I 0 mm/min. The values
for a"'"' are only reference values because of the test results depending on the
deformation rate and the length of the specimen. The properties of the fine-grained
concrete used for the specimens are given in Table 2.

Copyright American Concrete Institute

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
58 Hegger et al.
~ Previous tests [2], [3] showed that the tensile strength of the embedded fibers in
the composite material textile reinforced concrete cannot be fully exploited. Reasons for
this are the damage of the fibers during the textile manufacturing processes, the bond
characteristics of the rovings in the concrete, the fiber adjustment and the surface finish
of the fibers. For determining the tensile load bearing capacity of the composite material,
tensile tests were performed with 900 mm long and 100 mm wide test specimens as
shown in Fig. 4. Strains were measured directly on the specimens using LVDT's. The
loading was applied with a constant deformation rate of 1 mm/min. In order to examine
the influence of the fiber orientation, in one test-series the fabrics were turned around
22.5°, 45.0°, 67.5° and 90.0° with respect to the direction of the tensile stresses. In
addition, four-point-bending tests have been carried out to determine the load bearing
capacity of textile reinforced concrete structures under bending loading. The influence of
the reinforcement quantity on the load bearing capacity and on the effectiveness of the
fibers was examined by varying the reinforcement ratio. The specimens geometry and the
test set-up are shown in Fig. 5. The tests have been repeated two times, so that each test-
series with its specific material combination and test set-up consists of three tests.

Results: Investigations in [4] revealed that the shape of the textile reinforcement cross-
section has a substantial influence on the load bearing capacity. With a roving embedded
in the concrete, the filaments which are in direct contact with the concrete matrix transfer
higher bond forces than the filaments which are located inside a roving. Thus, rovings
with large diameters have worse load bearing characteristics compared to rovings with
small diameters, which have a more favorable cross-section area to perimeter ratio. This
is also confirmed by own tests as shown in Fig. 6, where the results of tensile tests are
compared. It is obvious that the fabric MAG-01-03 consisting of the finest ravings
reaches the highest utilization of the fibers. With increasing roving thickness, the
maximum failure stress of the textile reinforcement decreases.

The effect of the fiber orientation on the load bearing behavior of textile reinforced
concrete is shown in Fig. 7 (a). The load bearing capacity of sloped rovings is lower than
those of rovings aligned in the load direction. For the MAG-07-03 having an equal
orthogonal reinforcement the results show symmetry to an angle of 45°. The rate of the
loss of load bearing capacity subjected to the fiber orientation is given by the reducing
factor ko.a· With increasing fiber slope the effectiveness of the fibers decreases to 61% for
a fiber orientation of 45°. Reason for the loss of load bearing capacity are additional
stresses the sloped fibers are subjected to during the cracking process. The change of the
direction of the fibers at the crack edge causes bending stresses and delaminating of the
fibers from the matrix as well as a transverse force pushing the roving against the crack
edge. This leads to fiber failure at the sharp crack edge and can cause a failure at the
matrix edge. Tests in [5, 6] with sloped rovings showed that the fibers are not pulled out
of the matrix even if the bond length of the fibers is very short. In fact, the member's
failure is always caused by the fracture of the sloped fibers.

The four-point-bending tests on samples reinforced with MAG-07-03 showed that

there is no effect of the reinforcement ratio on the fiber effectiveness. In Fig. 7(b) the
reinforcement ratio is given as the percentage of reinforcement cross section area to the

Copyright American Concrete Institute

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
Thin Reinforced Cement-Based Products 59
concrete area of the profile. The fiber effectiveness k 0, shown as a function of the
reinforcement ratio, is calculated as:
- (J' f,bt
k() - (l)
(J' max

In this equation afht is the maximum failure stress of the roving in the bending test on the
textile reinforced concrete specimen and amax is the tensile strength of the roving. For
different reinforcement ratios the fiber effectiveness reaches a constant value of about
40% for the MAG-07-03. The fiber effectiveness may be improved by a coating. Coating
or laminating the textiles leads to the gluing of the filaments. Thereby, the bond
characteristics of the core filaments between each other are significantly improved and
the effectiveness of the fibers can be more than doubled [4].

Conclusions: Based on the test results, the tensile load bearing capacity Fc111 of the textile
reinforcement cross section embedded in concrete may be calculated as:
II

F;.,u = L kt; 'k~.a · A: '(J'1~ax {2)

i=l

where k 0 and ko.a are the factors accounting for the bond behavior and the orientation of
the fibers resulting from tension and bending tests on textile reinforced concrete
structures, A 1 is the cross-section area of the fabric, amax is the maximum tensile strength
of the ravings and n is the number of fabric types in the cross section.

DESJGN OF THE TEXTILE REINFORCED CONCRETE PANELS

The design of the textile reinforced concrete panels is based on the progress report for
textile reinforced concrete [2] and on the described experimental as well as theoretical
investigations on the behavior of textile reinforced concrete elements conducted at the
RWTH Aachen University.

The dimensions of the panels are 2685 x 325 x 25 mm and the support conditions are
shown in Fig. 8(a). Because of the statically determined support conditions no stresses
due to temperature changes are generated. The reinforcement layer in the longitudinal
direction lies about 4 mm from the surface of the panel leaving an effective depth d of 21
mm. This leaves a concrete cover of at least 3 mm. In addition to their own weight (g =
24 kN!m\ the panels are designed for a maximum wind suction Ws = 1.0 kN/m2
occurring at the comers of the building. The analysis resulted in a maximum bending
moment in the longitudinal direction of Mw.L. = 0.24 kNm/m due to wind loading as
shown in Fig. 9, in addition to a tensile force due to the own weight of the panels NnL =
2.8 kN/m. The design tensile force T, for the textile reinforcement is calculated as:
T = Yw.L.MwL + YDLNDL (3)
I
Yet 2

Copyright American Concrete Institute

60 Hegger et al.
Applying a moment arm Yet = 0.85d, a load factor 'YwL= 1.5 for wind load and 'YD.L. = 1.35
for own weight, results in a tensile force Tt = 22 kN/m. The required cross section At is
determined as:
YtT,
At = ---'--'-- (4)
kokoaumax

Considering the results of the described experimental investigations a coated alkali

resistant glass fiber fabric (2200 tex and 8.33 mm mesh size in longitudinal direction, and
320 tex and 8.4 mesh size in the transverse direction) has been chosen for the
reinforcement of the panels. The coating doubles the effectiveness k0 of the uncoated
fabric, thus, ko increased to 80 %. In addititon, using a coated fabric leads to a better
manageability because of the improved stability of the fabric structure. The fabric having
a maximum tensile strength amax of 627 MPa and including a material safety factor 'Yt =
1.5 for the textiles, the required area is 65.8 mm 2/m. The used mesh has a cross section
area of At= 97.4 mm 2/m in the longitudinal direction (direction of the tensile stresses).
The fabrics are arranged in two layers close to the surface, thus, providing an upper and
lower reinforcement layer. In the area of the bearings an additional layer has been
provided in the vertical direction as shown in Fig. 8(b). The load bearing capacity has
been checked in four-point-bending tests as shown in Fig. 10. The result of the tests was a
load bearing capacity of about 1, I 0 kNm/m of the panels (Fig. 11 ), meaning a maximum
tensile force of the textile reinforcement of 61.6 kN/m. Thus, the safety factor against

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
collapse under service loads is higher than 4. Furthermore, the large deflection values
measured in the tests indicate the ductility of the panels.

PRODUCTION OF THE TEXTILE REINFORCED CONCRETE PANELS

For the production of the panels a self-compacting fine-grained concrete having an

optimum consistency capable of fully soaking the fabric was used. The concrete mix used
is listed in Table 3. The panels were produced lying horizontally as shown in Fig. 12.
First, a 4 mm thick concrete layer is poured in the formwork, followed by the placing of
the first layer of textile reinforcement. Then a 17 mm thick layer of concrete is poured
followed by the placing of the second reinforcement layer, and finally the remaining 4
mm concrete layer is poured.

For fixing the curtain wall panels, an agraffe-fixing device shown in Fig. 13 is used.
The vertical aluminum substructure (skeleton) of the device is plugged into the steel-
reinforced wall. The agraffes are fixed to the textile-reinforced panels using special
dowels. These are positioned in the panel inside cone-shaped boreholes. Pull out and
shearing tests as shown in Fig. 14 have been carried out in order to check the load bearing
capacity of the dowels. In practice the dowels are loaded with a combined pull-out and
shearing load with a calculated maximum value of 0,17 kN. Therefore, the lowest load
capacity resulting from the tests had to be determined. The results (Fig. 15) showed that
the dowels can resist more than seven times the load they are actually subjected to in
practice even if they are positioned in cracked concrete.

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 61

PRODUCTION OF THE SANDWICH ELEMENTS

The concreting of the sandwich elements is performed in analogy to the production of

ordinary reinforced concrete elements. The textile reinforcement (in this case a contoured
profiled spacer fabric, consisting of two cover layers and threads in between as shown in
Fig. 16(a)) is placed in the formwork, which is then filled with highly liquid fine-grained
concrete. Before that, the anchors connecting the facing shell elements with the
reinforced concrete bearing elements are put in place. After the hardening of the facing
shells, the reinforced concrete bearing element is poured. In Fig. 16(b) a completed
sandwich element is shown.

SUMMARY

The successful application of textile reinforced concrete as exterior cladding panels in

a pilot project has been demonstrated. The thin walled, corrosion free panels proved to be
an efficient and economical alternative for conventional reinforced concrete or natural
stone facades. The new panels made of textile reinforced concrete open up new ways for
entirely new applications for concrete as a construction material.

ACKNOWLEDGEMENTS

The authors thank the Deutsche Forschungsgemeinschaft (DFG) in context of the

Collaborative Research Center 532 "Textile Reinforced Concrete" and the Ministry of
Labor and Social Affairs in the German State of North-Rhine/Westphalia (NRW) for
their financial support.

REFERENCES

I. Hegger, J.; Curbach, M., 2001, "First Colloquium on Special Research Areas 528 and
532", Proceedings, RWTH Aachen, Germany, February 15-16,334 pages

2. Curbach, M., Hegger, J. et al., 1998, "Sachstandbericht zum Einsatz von Textilien im
Massivbau", Chapter 7.5, Deutscher Ausschuss fUr Stahlbeton (DAfStb), Heft 488, Beuth
Verlag, pp. 81- 90

3. Ohno, S.; Hannant, D.J., 1994,"Modelling the Stress-Strain Response of Continuous

Fiber Reinforced Cement Composites", ACI Materials Journal, Vol. 91, No.3, pp. 306-
312

4. Molter, M., 2001, "Bruchtragverhalten textilbewehrter Biegekorper", Proceedings of

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

First Colloquium on Special Research Areas 528 and 532, RWTH Aachen, pp. 205-219

5. Mashima, M.; Hannant, D.J.; Keer, J.G., 1990, "Tensile Properties of Polypropylene
Reinforced Cement with Different Fiber Orientations", ACI Journal, Vol. 87, No. 2, pp.
172-178

Copyright American Concrete Institute

62 Hegger et al.
6. Bartos, P., 1982, "Bond in Glass Reinforced Cements in: Bond in Concrete", Elsevier
Applied Science, London, pp. 60- 72

NOTATIONS

A1 =cross-section area of the textile reinforcement

Fc1u = failure load from tensile test
k" ko =factors accounting for the bond behavior and the fiber orientation respectively
MwL. =bending moment due to wind loading
ND.L. =normal force due to own weight
T1 =design tensile force for the textile reinforcement
ul =perimeter of the textile reinforcement
Yet =internal moment arm
Omax =maximum tensile strength of the rovings
Yw.L. YDL =load factors for wind and dead loads respectively
y1 =material safety factor for the textile reinforcement
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

CONVERSION FACTORS

1 in. = 25.4 mm
1 ft = 0.3048 m
1 kip =4.448 kN
1 ft-kip = 1.356 kN-m
1 psi = 6.89x 1o-3 MPa

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 63

Table l - Fabrics used in tests

Name DirectiQn Roving Mesh Cross

thickness size section area strength

1240 4.2 110

1240 8.0 57

MAG-02- 0" 2480 8.3 110 553

03 90 1240 8.0 57 832

MAG-03- 0" 3100 8.3 138 445

03 90 1240 8.0 57 883

0" 2400 8.3 107 952

90 524
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

2400 8.4 106

Table 2- Concrete properties

Young's Modulus
(MPa)
33!00

Table 3 -Mix of fint>·grained concrete

Cement Fly Silica Plasticizer Sand Quartz Air w/c

Asn Fume 0-2mm flour
(ko/m)) (kg/m)) (kf!./m)) (kg/m 3) {kg/m 3) (kg/m 3) Vol.%
450 300 30 24.8 1178.5 539 0.5 0.5

Copyright American Concrete Institute

64 Hegger et al.
Facade Panels made of Textile Relinlf:•rcl!d

Existing Hall Extension

Fig. 1 - New extension of the testing hall of the Structural Concrete Institute
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

(a) Sandwich clcrncnis with facing shell of (b) Textile-reinforced cladding panels
textile concrete
Fig. 2 - Sandwich elements and cladding panels out of textile reinforced concrete

{a) View (b) Section

Fig. 3 - Curtain wall construction of the Structural Concrete Institute
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=University of Texas Revised Sub Account/5620001114, User=erur, ert
No reproduction or networking permitted without license from IHS Not for Resale, 01/26/2015 02:27:05 MST
Daneshlink.com
daneshlink.com

Thin Reinforced Cement-Based Products 65

LVOT

, 65 .,lr
'1
135 135 A' 65 >f
;f 200 400 200 ;f
900 [mm]

husked steel sheet

textile reinforcement

LVOT 0

II -4--ac- II
measurement range load introduction zone

Fig. 4 - Geometry and experimental arrangement of the tensile tests

"'"+•••••• •••·•••-~••••oom••••*•••oooo,.,•.,.,.,~,_., . .,. •• .,. •.,.,~ •.,. • .,.,.,~,., •.,., •., . .,.,..,,.,.
~ 1000 t-1

Fig. 5 - Test set-up of four-point-bending test and profile geometry

iii 600
0..
i!.
:a 500

~
'E 400

I
~
300

"£ 200

100

0
0,0 1,0 2,0 3,0 4,0 5,0 6,0
Strain (%o)
Fig. 6 - Reinforcement stress - strain of textile reinforced concrete element curves of
tensile tests
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

66 Hegger et al.

l::.~ J
";"
~M

I
J 0.2/ . . . ). . . .
0.2
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

....... .;... ...., ....:............ ~.........• 0

0,96% 1.44% 1.93%
0 15 30 45 60 15 90
Fiber oriontation M

a) Effect offiber orientation b) Fiber effecliveness kofrom bending rests

Fig. 7 - Reducing factors k0,3 and~ for MAG-07-03

50 I 168.5 I 50

L__ _ ;_·~:! !. .. .:. · ____·_·_Y_L_._x_____,!::!:!...··-···_ ___.j ~H .fl2.5

268,5
[cmJ
(a) Dimensions and support conditions of panels (in em)

(b) Amngcmcnt of textile reinfon::ement

Fig. 8 - Dimensions and reinforcement of textile reinforced concrete panels used

Fig. 9 - Bending moments in longitudinal direction of panels due to wind loading

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 67

L =:t:LVDT ~
25 tZ.c.;i.c:::::.:-::.:."'·•···:c~"··====~::::-i'~·n
'LVOT

Fig. 10 - Bending tests performed on textile reinforced concrete panels

0 20 40 60 80 100 120
Deflection [mm]

Fig. 11 - Results of bending tests performed on textile reinforced concrete panels

(al Rtinlbrccmcnt inside ll:•rmwork {b) concreting of dements

Fig. 12 - Production of the textile reinforced concrete panels

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

68 Hegger et al.
Vertical Substructure
(plugged at steel-
reinforced wall)

Agraffe

Dowel in
cone-shaped

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
borehole

Panel
Fig. 13 - Fixing technique of the curtain wall panels

(a) shearing test (b) pullout test

Fig. 14 - Tests performed on the fixing dowels

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 69

2,99
Interpolation
/
/,._...r

Sheariogia<><l 2, 78 Shearing load 2,11!

bearing C<lf'<1City {kNl bearing r;apaci1y [i<N]

(a) fixing dowel in cmckcd concrete (b) thing dowel in uncrncked

concrett'
Fig. 15 - Load bearing capacity of the fixing dowels

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
{a) profiled spacer tabric

(b) c~lmpkt<'d sandwich clement

Fig. 16 - Sandwich elements used

Copyright American Concrete Institute

70 Hegger et al.

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

SP-224-06

Ultra-High Performace Concrete with

Ductility: Design, Prototyping, and
Manufacturing of Panels and Boxes

by D. Zakariasen and V. Perry

Synopsis: Ductal'" is a new material technology offering a unique combination of superior

characteristics including ductility, strength, and durability, while providing highly
moldable products with a quality surface. The technology provides compressive
strengths up to 200 MPa (30,000 psi), and flexural strengths up to 50 MPa (7 ,200 psi).

The material's unique combination of superior properties enables the designer to

create thinner sections, longer spans, and higher structures that are lighter, more graceful
and innovative in geometry and form while providing superior durability and impermeability
against corrosion, abrasion, and impact. This material provides the precast industry with
opportunities to improve many existing products and manufacture new products that will
compete with other materials such as stainless steel, cast iron, ceramics, and others.

This paper presents properties of the material, design assumptions for project
solutions and the manufacture, installation and assembly procedures for specific projects
including roof panels, 5 sided-boxes and anchor plates.

Many economies gained from this new technology are a result of engineering new
solutions for old problems. By utilizing the unique combination of superior properties, designs
can eliminate passive reinforcing steel and experience reduced global construction costs,
form works, labour and maintenance. Additionally, this relates to benefits such as improved
construction safety, speed of construction, extended usage life and others.

Keywords: abrasion; aesthetics; composite; ductile; durability; fiber-

reinforced; impact; impermeability; UHPC; usage-life

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

71
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=University of Texas Revised Sub Account/5620001114, User=erur, ert
No reproduction or networking permitted without license from IHS Not for Resale, 01/26/2015 02:27:05 MST
Daneshlink.com
daneshlink.com

72 Zakariasen and Perry

BIOGRAPHIES

V.H. (Vic) Perry, FCSCE, MASc., P. Eng., Vice President and General Manager -
Ductal®, Lafarge North America Inc. (Calgary), received his Bachelor of Civil
Engineering (1978) and MASc in Structural Engineering (1984) from Dalhousie
University, Halifax, Canada. Mr. Perry completed Executive Management Programs and
Graduate Studies at the University of Western Ontario, University of Toronto and Duke
University. Mr. Perry has been involved with Ductal® since 1997, initially as Marketing
Director- Ductal®, for the Lafarge Group in Paris, France.

Don Zakariasen, Manager - Business Development for Lafarge Canada Inc./Precast

(Calgary), has 32 years experience in the precast/prestressed related industry, is the
immediate Past-Chairman of the Canadian Precast Concrete Institute and served on the
Board of Directors for the Precast Concrete Institute in the U.S.A. Prior to Lafarge, Mr.
Zakariasen was President of Bonnybrook Custom Steel Forms, a major supplier of steel

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
form work to the North American precast industry.

INTRODUCTION

Ductal® is a revolutionary, ultra-high performance material marketed in North

America by Lafarge North America Inc. This new material technology has a unique
combination of superior technical characteristics including ductility, strength and
durability, while providing highly moldable products with a high quality surface aspect.
It provides compressive strengths of up to 200 MPa (30,000 psi), with flexural strengths
of up to 50 MPa (7200 psi). The products made from this material can reduce global
construction costs and labour requirements, improve construction safety, lower
maintenance and increase a structure's life span.

The unique combination of superior properties enables the designer to create

thinner sections, longer spans and higher structures that are lighter, more graceful and
innovative in geometry and form while, at the same time, providing superior durability,
impermeability, and resistance against corrosion, abrasion and impact. It provides the
precast industry with opportunities to improve many existing products and manufacture
new lines of products that will compete with other materials such as stainless steel, cast
iron, ceramics and others.

Many economies gained from this new technology are a result of engineering
new solutions for old problems. By utilizing the unique combination of superior
properties, designs can eliminate passive reinforcing steel and experience reduced global
construction costs, form works, labour and maintenance. Additionally, this relates to
benefits such as improved construction safety, speed of construction, extended usage life
and others.

This paper presents properties of the material, design assumptions for project
solutions, manufacture, installation and assembly procedures for specific projects. The
projects described include roof panels, 5 sided-boxes and anchor plates.

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 73

DUCTAL® MATERIAL CHARACTERISTICS

Ductal® is a new material technology with a unique combination of superior

technical characteristics including ductility, strength and durability, while providing
highly moldable products with a high quality surface aspect. It is a high-strength ductile
material formulated of constituent materials. These materials include portland cement,
silica fume, quartz flour, fine silica sand, high-range-water reducer, water and steel or
organic fibers. The technology of the material is covered by one of many patentse) in a
range of Ultra-high Performance Concretes, all under the Ductal® trademark, providing
compressive strengths of up to 200 MPa (30,000 psi) and flexural strengths up to 50 MPa
(7,200 psi)(\

The ductile behaviour of this material is a first for concrete. The material has
the capacity to deform and support flexural and tensile loads, even after initial cracking
(Figure 1). These performances are the result of improved micro-structural properties of
the mineral matrix, especially toughness and control of the bond between the matrix and
the fiber.

There is almost no carbonation or penetration of chlorides and sulphides and a

high resistance to acid attack. The superior durability characteristics are due to a
combination of fine powders, selected for their relative grain size (maximum 600

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
microns) and chemical reactivity. The net effect is a maximum compactness and a small,
disconnected pore structure.

The material has almost no shrinkage or creep, which makes the material very
suitable for prestressed applications. The use of this material for construction is
simplified through the elimination of reinforcing steel and the ability ofthe material to be
virtually self-placing or dry-cast.

The following is an example of the material characteristics published by the

supplier(\

STRENGTH e)
Compressive (3" x 6" cylinders) 160- 220 MPa (23- 33 KSI)
Flexural (I Yz" x I Yz" x 8" Prism) 35-50 MPa (5000- 7200 PSI)
Youngs Modulus (E) 55-60 Gpa (8- 8.5 x 106 PSI)
Total Fracture Energy 20,000 - 30,000 11m 2 ( 1300- 20001b (F)-ft/ft2)
Elastic Fracture Energy 20-30 J/m 2 2
(1.3 -2.0 lb (F)-ft/ft )

DllRABIUTY e)
Chloride ion diffusion (CI) 0.02 X ] 0" 12m 2/s (0.02 x 10" 11 ft 2/s)
Carbonation penetration depth <0.5 mm (<0.02 inches)
Freeze/thaw (after 300 cycles) ASTM 672 100%
Salt-scaling (loss of residue) ASTM 666 <10 g/m 2 ( <0.0025 lb/ft2)
Abrasion (relative volume loss index) 1.2

Copyright American Concrete Institute

74 Zakariasen and Perry

The materials are supplied to the precaster in a three component pre-mix. The
powders are pre-blended in 'bulk-bags' at a blending facility and shipped to the precast
plant. The superplasticizer and steel fibers are supplied separately and introduced into
the precaster's mixer at the time ofbatching.

MARKETS FOR THIS TECHNOLOGY

Ductal® is a new material with a unique combination of technical characteristics

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

unlike any other material currently available. Using these combination of properties to
design new solutions for old problems is 'key' to providing ecomonical solutions to
customers. A few of the markets that have been explored and appear interesting are:

Long-span Roofs - spans greater than 20 m (50'), particularly in combination with

abrasion, corrosion or heavy loads.

Anchor Blocks/Plates - ability to resist heavily concentrated loads in corrosive

environments, with enhanced aesthetics.

Kennels/ Animal Proof Waste Containers- thin walled, aesthetically pleasing, corrosion
resistant and durable against denting.

Acoustic Pannels - thin panels replicating gypsum acoustic panels capable of resisting
corrosion and impact or abuse.

Structural Wall Panels - thin-walled with rib, light weight load bearing panels, with
enhanced architectural finishes.

Bridges - long spans, shallower profiles, lighter weight spans with improved
impermeability, abrasion resistance, durability.

EXAMPLE OF PROJECTS

The following 1s a selection of projects which demonstrate the use of this

technology.

The Joppa Project

The Joppa project is an upgrade to an existing manufacturing facility in Joppa,

Illinois, USA. The total upgrade, at an estimated cost of twenty million US dollars,
included the installation of three new clinker silos. It is the roofs of the silos that were of
interest for a Ductal® versus steel analysis of a long..:span roof.

This technology was used for one of the three clinker silo roofs (Figure II) and a
conventional steel solution for the other two. This section presents the properties of the
material, the design assumptions for the two roof solutions, the manufacture of the panels
and the erection and assembly of the roofs.

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 75

The clinker silo walls are slip formed concrete of 18 meters (58'), 30 meters
(100') and 37 meters (121') in diameter with truncated, cone shaped roofs of Ductal® or
steel. The roofs are attached to the top of the slip form silo walls which provide a
tension ring for the base of the roof. The top of each roof supports a mechanical
penthouse floor with an intregal steel compression ring attached to the roof.

The Ductal® roof consists of twenty-four precast, pie-shaped panels with a 15

mm C!8 ") skin thickness for an 18 m (58') diameter silo roof (Figure III). The panels
were designed to act as a thin shell, supporting a two-storey mechanical penthouse
centered at the top of the cone shaped roof. The ultra-light, thin precast panels did not
use any reinforcing bars.

Due to fast track requirements and the owner's concern that a new technology
may interfere with the schedule, a decision was made to carry two designs in parallel
through to the award. Both options were designed by the Engineers and tendered
competitively.

Table I summarizes the test data results from the production of the 24 roof
panels cast at the Winnipeg precast plant. For the roof panels, "Ductal® CS I 000" was
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

specified in the contract document. Ductal CS I 000, one of several in the Ductal® product
range, is formulated for structural use in civil engineering projects. The mean
compressive strength of 166 MPa (24,000 psi) was supplied to comply with the
specification minimum requirement of 160 MPa (23,000 psi).

The Design -- The unique combination of superior technical characteristics

means that using the material in a conventional manner will not maximize the efficiency
of the material. Therefore a new solution was required for the Joppa roof.

Considering that (today) Ductal® is a precast product and the owner's

requirements for an airtight solution, several shell options were considered. Taking into
consideration these constraints plus the loading, and geometric requirements and
properties of the material, a thin shell plate was selected. It was also decided that a
truncated cone would be the most efficient roof profile.

From this point, finite element analysis was used to predict the shell stresses and
potential buckling. The principal compressive stresses due to the conical forces resulted
in a required shell thickness of I 0 mm (I /2" ). Buckling, demolding, handling,
transportation and erection conditions dictated radial side stiffener beams, pre-stressed
with four strands, at each side of the precast panels. Secondary bending stresses due to
uniform roof live loading and buckling requirements dictated transverse ribs. See Figure
IV for a plan and cross-section of a typical panel.

The final finite element modeling of the roof with all loads and geometry
revealed very low service stress levels. Minimum cross-sectional dimensions and
buckling generally governed the final design.

Copyright American Concrete Institute

76 Zakariasen and Perry

The steel roof (Figure V) was a conventional steel frame with radial beams and
purlins covered with a corrugated steel deck membrane.

Manufacturing & Transportation of the Panels - The precast panels were

manufactured (Figure VI) at a plant in Winnipeg, Manitoba, Canada and trucked to the
site in Joppa, 111inois, USA. A three-component pre-mix was supplied to the precast
plant and hatched in a two cubic meter Nikko horizontal twin turbine mixer.

Batch sizes of0.9 m 3 (1.2 yd 3) were mixed and placed into the steel forms in one
continuous casting. QA/QC testing on each batch included flow tests, fiber content,
cylinders for compressive strengths and prisms for flexural strengths.

Following casting, the panels were covered with a tarpaulin and steam cured at
40°C (IOOF) for 16 hours. The following day, when the panels attained 40 MPa (6000
psi) minimum compressive strength, the strands were cut and the panels removed from
the forms. At the end of each week, the panels were stacked in piles of 4 and thermally
treated at 90 °C ( 190F) for 48 hours (Figure VII).

Transportation of the panels from the Winnipeg, Canada precast plant to the
Joppa, USA site was via flat deck trucks in loads of eight panels per truck (three loads for
a total of twenty-four panels).

Erection and Installation of the Panels - The panels arrived on site in three
truck loads of eight panels per truck (Figure VJII). All twenty-four panels were unloaded
and staged in position for installation to the roof.

A mobile crane was used to install and temporarily hold the circular steel beam
penthouse floor in place until the panels were installed. The panels were lifted one at a
time (Figure IX), alternating opposite sides and bolted to the steel penthouse floor and the
top of the concrete silo wall (Figure X).

When all panels were bolted into place, the crane was disconnected from the
penthouse floor and freed up to start other work on the site. Grouting between the panels
completed the air and water tight roof system. The penthouse and connecting conveyors
were then installed (Figure XI).

The Ductal® roof system resulted in significant on site construction time savings
(Table II). It took eleven days of construction to install the Ductal® roof compared to 35
days for the steel roof system. Additionally, it was discovered that the roof panels were
more accommodating to the construction tolerances for out-of-roundness and flatness of
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

the top of the slip-formed silo walls. Whereas the steel roof elements were premanu-
factured to exact tolerances and required additional closure plates and redrilling to install
on the out-of-round silo walls, the precast panels were supplied with a roof overhang and
over-sized holes to receive expoxy pins to connect to the top of the silo walls.

Copyright American Concrete Institute

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
Thin Reinforced Cement-Based Products 77
This project (Figure XII) was the first of its type in the world for the use of
Ductal® in a long-span roof structure. While this solution demonstrates many of the
benefits of the technology, it is apparent that the true benefits of are not yet fully
recognized. Furthermore, the optimized profiles and use of this technology is in its
infancy and, in the next few years, much progress is anticipated in the area of optimized
solutions using this material.

Animal Proof Waste Containers & Dog Kennels

Animal proof waste containers and dog kennels are both thin-walled, 5-sided
boxes, with doors/gates- one designed to keep animals in and the other designed to keep
animals out. Both products require a high quality surface finish on all surfaces with
rounded comers to prevent injury. The kennels and waste containers are required to
resist denting, rusting and abuse in a corrosive environment.

The kennels are 1.8 m x 1.5 m x 1.2 m (6' x 5' x 4') with a wall thickness of 20
mm (3/4"), for a total material volume of 0.2 m 3 (0.26 yd 3) per kennel. The waste
containers are 1.0 m x 0.5m x 0.5 m (3' x 18" x 18") with a wall thickness of 20 mm
(3/4"), for a total material volume of0.05 m 3 (0.06 yd 3) per container.

The selected production method was to use double sided molds for a 5-sided box
(Figure XIII) and to use an injection casting technique (Figure XIV). All elements were
cast upside down and injected at the bottom of the mold under a constant pressure. A
pressure vessel filled with the fluid material was connected to the mold with a flexible
tube through a knife valve gate. Typical injection casting times to fill the mold were I 0
minutes. The quality and tightness of the molds were very important to ensure a high
quality surface finish and to not have leakage of the fluid material while injecting.

The final products (Figures XV & XVI) were an excellent example of new
products with improved charact~;ristics and economical benefits.

Anchor Plates I Blocks

In 2002, precast retaining wall panels were manufactured for placement beneath
a highway bridge in Calgary, Alberta, Canada. Due to restricted site accessibility, it was
decided that off-site construction of the precast wall panels would speed up the schedule
by permitting earlier installation of the super-structure prior to completing the
earthworks, sub-structure and retaining structures.

The engineer designed a post-tensioned, soil anchor precast retammg wall

system, bearing on a lean concrete strip footing, utilizing soil anchor rods grouted into
stable soil. The soil anchor rods were then bolted with anchor plates into pockets, cast
into the precast wall. The anchor plates were designed (in parallel) in both galvanized
steel and Ductal.g,.

Copyright American Concrete Institute

78 Zakariasen and Perry

The retaining wall consisted of 42 precast units of 4.9 meters (16') in length,
varying in thickness from 300 mm to 550 mm (12" to 22") and up to 3.1 meters (1 0'-2")
in height (Figure XVII). Each precast wall panel is set in place on a lean' concrete mud
slab, then tied back with steel soil anchors to resist the soil pressure (Figure XVIII).
Following the grouting of the anchor rods into stable soil, 64 Ductal® anchor blocks were
installed and a bell nut threaded onto the end of the rod.
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

The anchor blocks, manufactured at Lafarge's precast plant in Calgary, were

506 mm x. 506 mm x 280 mm thick (20" x 20" x 11 ") and used approximately 4m 3 (5 yd 3)
of material for the entire project.

Steel molds were manufactured for casting the block-outs in the precast
retaining walls. The same molds were then used for the anchor blocks, resulting in a
match casting of the blocks and pockets (Figure XIX). This also provided molds with
water tight joints to prevent leakage of the fluid during the casting operations. A plastic
insert was cast into the top side of the mold to form a precise pocket to accept the counter
sunk bell nut.

The specifications required a compressive strength of 150 MPa (22,000 psi) and
the mean compressive strength obtained was 188 MPa (27 ,000 psi). Batch sizes were
0.22 m 3 (0.3 yd 3 ) to cast 3 anchor blocks in one casting. Total production took 11 days.
The anchor blocks were assembled in the precast yard on pallets (Figure XX) and shipped
to the project site as required by the schedule. Each block had a mass of 160 kg (350
lbs).

The anchor blocks were installed at the site (Figure XXI) using a bobcat and two
men. Following the placement of the blocks into the wall pockets, the bell nut was
threaded onto the soil rod (Figure XXII & XXIII).

CONCLUSION

Ductal's® unique combination of superior properties enabled the designer to

create thinner sections, longer spans for a tall structure that is lighter, more graceful and
innovative in geometry and form. Overall, solutions with this material offer advantages
such as speed of construction, improved aesthetics, superior durability, impermeability,
and resistance against corrosion, abrasion and impact - which translates to reduced
maintenance and a longer life span for the structure.

REFERENCES

I. Patents: Issued by the U.S. Patent Office 5, 503, 670 and, 5, 522, 926.

2. La farge North America Inc., Technical Characteristics sheet for Ductal® with
Metallic Fibers, found on the website: www.imagineductal.com.

3. Perry, Vic, "HPC Bridge Views", Q&A, Issue No. 16, July/August 2001.

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 79

4. Strudes Consulting Engineers, "Joppa Project Drawings and Specifications",
Project No. 996979, Joppa, II, USA, August 2000.

5. AFGC(Association Francaise de Genie Civil) Interim Recommendations

"Ultra High Performance Fiber-Reinforced Concretes", January 2002.

6. J.R. Spronken & Associates Limited, "Country Hills Retaining Wall (Ductal®
Anchor Blocks) Project Drawings and Specifications", Project No. COI098,
Calgary, AB, Canada, March 2002.

7. Mouloud Behloul. "Analyse et modelisation du comportement d'un materiau a

matrice cimentaire fibree a ultra hautes performancces (BPR). Du materiau a
Ia structure". These de Doctorat. Ecole Normale Superirure de Cachan. 13
December 1996.

T able :Ductal-"'Test Resu ts- J oppa Precast P ane s

Property Mean Value Mean Value After Std Dev.
@ 20 hours - Release 48 hrs Thermal
Treatment
Compressive 50.0 MPa 166.0MPa 10.8 MPa
Strength (7 ,250_l)Si) (24,000 psi) (1.500 psi)
Flexural - 37.9 MPa 4.1 MPa
Stren"th (5,500psi) (600 psi)

Ta ble 11 : c ompanson of the Roo fl nsta 11 at10n

° sc b ed ule
Item Steel Roof Ductal' Roof
(davs) (days)
Pre-Erection Survev 2 I
Pre-erection set-up (cranes & scaffolding) 5 I
Installation of Penthouse Floor I I
(Temporary Compression rin_g)
Installation of Ductal" panels - 3
Grouting of Ductal'' Panels - 5
Installation of Steel Roof Framing 12 -
Installation of Steel Roof Deck (Membrane) 10 -
Grouting Steel Base Plates for beams at top of silo walls 1.5
Installation of Closure Ring & Flashing for 3.5 -
Steel Roof Deck (Membrane)
Total (Construction davs) 35 It

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

80 Zakariasen and Perry

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
1.0 1.5
Deflection (mm)

Figure I - Load deflection curve for Ductal prism

Figure II - Schematic: General arrangements -clinker storage & handling

Plan 24 Precast Ductal Roof Pa:nels ..,

- 1.95 metric tonnes each (430QJh$r •
-I U m: !panel 029,s;q:ft::i"
,.., ......

........... ...
.. -........ ~

Figure III - Plan and section of the clinker storage silo

Copyright American Concrete Institute

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
Thin Reinforced Cement-Based Products 81

Figure IV - Plan and cross-section of a typical panel

Figure V -Steel framing (left) vs. Ductal framing (right) for the silo roof

Figure VI - Casting of a panel at the precast plant

Copyright American Concrete Institute

82 Zakariasen and Perry

Figure VII - Stack casting of panels for the 48 hr thennal treatment at 90° C

Figure VIII -Trucking eight panels per truck.

Figure IX - Site staging and erection of the panels

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 83

Figure X - Flying Ductal panels into place using a mobile crane

Figure XI -View from underside. looking up at Ductal panels

Figure XII- All 3 clinker silos and connecting feeder conveyors. (Ductal silo at far right).
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

84 --``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
Zakariasen and Perry

Figure XIII - Inside of mold, showing openings for the top and rear doors

Figure XIV- Set-up for injection casting of animal proof waste containers.

Figure XV - Form and prototype waste container.

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 85

Figure XVI - Six finished dog kennels.

1700mm
.
:. .r--
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Figure XVII - Schematic of retaining wall with Ductal@ anchor blocks.

Figure XVIII- Precast retaining wall.

Copyright American Concrete Institute

86 --``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
Zakariasen and Perry

Figure XIX - Casting the anchor blocks.

Figure XX -Anchor blocks ready for shipping

Figure XXI - Site visit during installation of anchor plates.

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 87

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Figure XXII -Anchor plate installed in the precast retaining wall.

Figure XXIII -Anchor plate installed, elevation view.

Copyright American Concrete Institute

88 Zakariasen and Perry

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

SP-224-07
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

New Cement Composites for Thin

Structural Products

by E. Parant and P. Rossi

Synopsis: This paper first proposes a reviewing and a critical analysis of the different
UHPFRC which exist, and secondly presents a new cement composite, the
CEMTEC,, 111"'"".,. patented by the Laboratoire Central des Pants et Chaussees (Paris,
France).
This cement composite has been tested under static bending and asymmetric fatigue
bending From this experimental study, the following comments can be made :
The characteristic strength and ultimate strain in compression are respectively
equal to 205 .MPa, and 4 J0· 3 .
the J-oung modulus is equal to 55 GPa and the Poisson coefficient is equal to
0.21.
The average modulus of rupture (MOR) is equal to 61.5 MPa;
n1e average strain related to the average MORis equal to 9.2 1(}3 •
A critical initial static strain threshold exists. Before this threshold a specimen
in CEMTEC,111,;,c.de'' does not fail during a bending fatigue loading and beyond
this threshold the failure fatigue cycles number linearly depends of the initial
static strain. The strain threshold determined in this study is befv..!een 1.24 x 1&
3
and 1.44 x J0· 3.
Below a loading ratio R = 0.65,failure during bendingfatigue test never
appears with a specimen of CEMTEC'",,,;_,cule G!·.

Keywords: cement composite; fatigue behavior; multiscale fiber

reinforcement; static behavior

89
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=University of Texas Revised Sub Account/5620001114, User=erur, ert
No reproduction or networking permitted without license from IHS Not for Resale, 01/26/2015 02:27:05 MST
Daneshlink.com
daneshlink.com

90 Parant and Rossi

Pierre Rossi is research director and head of the Concretes and Cement Composites
department at the Laboratoire Central des Ponts et Chaussees (LCPC) in Paris (France).
He is member of ACI committee 544, Fiber Reinforced Concrete. His research interests
are related to modeling of cracking of concrete structures including delayed and impact
aspects, Steel Fibre Reinforced Concretes and High Performance Concretes.

Edouard Parant is a doctoral candidate in the Concretes and Cement Composites

Department of the Laboratoire Central des Ponts et Chaussees, Paris, France. His
research interests include mixing procedure, mechanical characterization, fatigue and
impact loadings, and corrosion under stress of multi-scale cement composites.

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
INTRODUCTION

The recent development history of Fibre Reinforced Cement Composites has been
marked by three separate approaches. In chronological order of their appearance they are:

- FRCC containing between 5 and 10% of metal fibres which are 6 mm in length and
0.15 mm in diameter. This type of concrete was developed by the company Aalborg
Portland (Denmark) and has been marketed under the name CRC (Compact Reinforced
Composites) [1 ];

- FFRC containing a maximum of 2.5% of metal fibres which are 13 mm in length and
0.16 mm in diameter. This type of concrete was developed by Bouygues, Lafarge, and
Rhodia which are french companies (France) and has been marketed under the name of
DUCTAL~ [2];

- FFRC containing mixtures of short and long metal fibres. This type of concrete, called
MSCC (Multi-Scale Cement Composite), was developed by the Laboratoire Central des
Ponts et Chaussees (LCPC, France) [3).

Two types of fibre reinforced concrete have been developed concurrently with those
described above, but we feel they differ slightly from them. The reason is that these two
materials do not possess an ultra-compact matrix, i.e. a matrix with ultra-high
compressive strength (~ 150 MPa) as their matrix compressive strength does not exceed
70 MPa. They are mainly of interest because under uniaxial tension they are ductile and
exhibit strain-hardening and therefore undergo multicracking under tension.

SIFCON (Slurry Infiltrated Fibered Concrete) [4] is produced by filling the formwork
with bulk fibre and then injecting a fluid mortar slurry which coats the fibres. This
technique results in a very high fibre content of between 7 and 15% depending on the
shape and the laying technique. Consequently, to achieve correct filling, the slurry must
be extremely liquid which means that water-cement ratios must be very much higher than
for other types of FRC.

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 91

ECC (Engineered Cementitious Composite) [5] is produced by using synthetic fibres with
a high Young's modulus. These fibres are generally not more than 20 mm in length and
their diameter is usually less than 0.05 mm.

- CRC · The Danish decision to use a high percentage of short fibres (between 5% and
I 0%) can lead to an improvement in the mechanical characteristics of the material under
tension (strength and ductility) but cannot have a significant effect on the bearing
capacity and ductility of a structure, with the exception of very thin structural members
(because of the scale effects). Consequently, for thicker structural members such as slabs
or beams, CRC is used with a very high percentage of conventional concrete
reinforcement (5 or I 0 times the usual percentage). This combination of a high
percentage of metal fibres and a high percentage of conventional reinforcement is logical
and mechanically effective for the following reasons :

- As the high percentage of short fibres increases the tensile strength of the material more
than its ductility, in order to construct a ductile structure with this type of concrete, the
use of a high percentage of reinforcement is imperative,

- We know that structures are weakened when a certain percentage of reinforcement is

exceeded, firstly because very high stress concentrations are created at the surface of
the structure leading to a large number of cracks, and secondly because the reinforcing
bars are subject to an adverse group effect which causes them to operate individually
much less effectively (there is more cracking around the bars). Short fibres help to
control cracking at the surface and around reinforcing bars.

This combination of two different types of reinforcement which operate at different

scales is therefore excellent.

However, CRC does have a number of shortcomings, the main ones being :

· It is an extremely costly technical solution both as regards material and labour (the
reinforcing bars are difficult and time-consuming to install);

- 1t restricts freedom as regards the shape of structures.

- DUCTAL '!l - This material uses fibres which are twice as long as those in CRC but with
the same diameter. The choice of fibres with a very high specific surface area has two
consequences :

- The first is that to achieve the same level of workability obtained for CRC with between
5 or I 0% of fibres, it is barely possible to exceed a fibre content of 2.5% with
DUCT ALR. This relatively low fibre content has two adverse effects on the mechanical
performance of DUCT AL'1j;: .

the fibres do not raise the uniaxial tensile strength of the matrix which is about 8 Mpa ;

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

92 Parant and Rossi

in such a fragile matrix, as there is not a sufficiently high percentage of fibres, the post-
cracking behaviour of the material is highly variable.

This shows that by adding a single size of metal fibres to an ultra-high strength matrix it
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

is not possible to achieve effective action at the scales of the material and the structure:

- Bonding between the fibre and the matrix (CRC and BPR have a similar matrix) is
considerably better in the case of BPR than CRC. The fibres can therefore have an
effect on wider cracks, and therefore improve the performance of the structure (in terms
of bearing capacity and ductility). Unlike CRC, there is no need to use very large
amounts of conventional reinforcement in order to obtain ductile structures.

The following remarks can be made concerning the scope for industrial applications of
DUCTAL®:

- In view of its relatively low direct tensile strength (10 MPa), DUCTAL® can only be
used in reinforced or prestressed concrete beams or slabs to replace transverse
reinforcement (for example to resist shear force in frames). Furthermore the high cost
of the material means that it is essential that the specification for the structure makes
great demands as regards to durability and lightness ;

- The mechanical performance scattering of FRCC depends on the percentage of fibres

which are correctly oriented with respect to the cracks. With DUCTAL®' as the
percentage of fibres in relation to the strength of the matrix is not very high, it is
essential to orient the fibres correctly with respect to the cracks which will appear in the
structure. In the case of thin structural members, this requirement will be met, as the
fibres are necessarily oriented orthotropically, perpendicular to flexural cracks (such
members are essentially subjected to flexural stresses). It is therefore possible to obtain
thin DUCTAL® members that are mechanically homogeneous and ductile.

- MSCC- These materials are a direct application of the Multi-Scale Fiber Reinforcement
Concept [6]. The idea is to mix short fibres with longer ones in order to act both at the
scale of the material (increasing tensile strength) and the scale of the structure
(increasing bearing capacity and ductility).

Using this approach, there is no difficulty involved (for achieving good workability) in
adding, for example, a percentage of fibres that may be as high in volume terms as 7%.
With respect to the uniaxial tensile performance, the material exhibits a stress-hardening
behaviour (in the same way as steel) and strengths of 15 MPa are readily achievable. The
MSCC in question consisted of a mixture of 5% of straight drawn steel fibres, 5 mm in
length and 0.25 mm in diameter, and 2% of hooked end drawn steel fibres which were 25
mm long with a diameter of 0.3 mm.

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 93

A NEW CEMENT COMPOSITE: CEMTECmulti<cale®

CEMTECmultiscale ~. which was the subject of a world patent filling by the LCPC in March
2001, is conceived starting from the same concept as a MSCC, but with some evolutions
compared to this last. These evolutions are declined as follows :

- whereas the MSCC contains 2 different metal fiber geometries, CEMTECmlllti<cale®

contains 3 of them ;

- CEMTECmultiscale <& contains 11% per volume of fibers whereas the MSCC contains 7%
of them.
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

The LCPC launched, in 2000, a vast study over 4 years on CEMTECmultiscale®• study
which comprises mechanical tests to characterize the various mechanical behaviors of the
composite (behavior in statics, in fatigue, at high strain rates ... ), tests of durability, tests
on structural elements, and tests to optimize the manufacturing process (mixing and
casting). In this article only the results related to static and fatigue tests are presented.

The composition ofCEMTECm,dtiscule® is given in table 1.

STATIC TESTS

One of the industrial application aimed with this new composite material relates to the
slabs and the tloors strongly charged, such as the slabs of composite structures for
example.

To dimension a structure or a structural element made up only of CEMTECmultiscale®

necessarily implies to do it by taking into account in a very controlled way of safety. It is
consequently essential to reach characteristic mechanical behaviors which integrate the
dispersion problems inherent in all materials. It is thus, that the tests evoked above were
carried out on a sufficient number of specimens to determine these characteristic
behaviors. The static tests presented in this study aim on one hand at determining the
characteristic bending behavior and on the other hand the characteristic compressive
behavior of CEMTEC.wttiscale<R within the slab. In order to optimize the dimensions of the
specimens with respect to the scale effects and of preferential orientation of fibers, and in
addition to lead to a use which could be economically viable of CEMTECmultiscale®, it was
selected to retain following dimensions concerning the specimens representative of
a slab:

length: 600 mm,

width: 200 mm ,
thickness : 40 mm.

The 200 mm width allows an orthotropic orientation ofthe fibers, of which largest makes
25 mm length, representative of this existing in a slab.

Copyright American Concrete Institute

94 Parant and Rossi

Concerning the castings, the specimens representative of a slab are cast flat and are
vibrated during the casting on a mobile plate; whereas the specimens related to the
compressive tests are obtained by sawing, within additional specimens related to the
bending tests in order to have a realistic orientation of fibers compared to that which
exists in the compressive zone of the slab. The specimens for compressive tests are thus
prismatic.

It is known that with the Ultra High Cement Composites, which is CEMTECmultiscate®• the
use of a heat treatment makes it possible to increase the mechanical performances of the
matrix. Also, in the present study it was used a heat treatment which consists in placing
the specimens in a drying oven at 90°c during 4 days, 48 hours after their release from the
mould.

Lastly, it also should be announced that, parallel to the compressive tests on the prismatic
specimens, compressive tests were also realized on cast cylindrical specimens. Thus, 6
cylindrical specimens with a diameter II em and height 22 em were used to determine
the compressive strength, and 3 cylindrical specimens with a diameter 16 em and height
32 em were used to determine the Young modulus and the Poisson coefficient. All the
cylindrical specimens were thermically treated.

Four points bending tests- test set up

During these deflection tests the distance between the lower supports is 420 mm, and
between the higher supports is 140 mm. The test is carried out at an imposed deflection
rate equal to 0,3 mm/min. The deflection is measured using a special extensometer,
placed on the specimens, designed to eliminate parasitic displacements on the level from
the supports.
9 specimens were tested.
lt is also necessary to announce that for all the specimens, a L VDT sensor was stuck on
the face opposed to that on which the extensometer being used is fixed to measure the
deflection. This sensor is positioned at the level of the bottom fiber of the specimens in
the zone of constant moment.

Compressive behavior- results

The principal results related to the compressive behavior of CEMTECmultiscale® are the
following:

- The average strength and ultimate strain in compression are respectively equal to 220
MPa, and 4.5 I o-3•
- The characteristic strength and ultimate strain in compression are respectively equal to
205 MPa, and 4 I o-3 .
- The Young modulus is equal to 55 GPa and the Poisson coefficient is equal to 0.2I.

Bending tests- Results

In figure 1 are presented the results in the shape of bending tensile stress-strain curves.

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 95

One can note that the average bending tensile strength, still called average modulus oj
rupture (MOR), is very high. It reachs 61.5 Mpa;
The average strain related to the average MORis equal to 9.2 10-3 .

FA TIGUE TESTS

15 specimens were tested in bending fatigue, in an imposed loading rate (same test set-up
than for static tests). A first slow loading rate is imposed to reach the chosen maximal
fatigue load, then the load is decreased to reach the average fatigue load, and finally the
asymmetric sinusoidal fatigue loading is taken between I 0 and I 00 % of the chosen
maximal fatigue load. The imposed loading frequency was of 2.5 Hz. The test room is a
20 oc air-conditioned room. Tests were led until 2 millions cycles (10 days), except in
the case of premature failure.
During the study, the loading ratio R (ratio between the applied stress and the
characteristic static stress) varied between 0.77 and 1.03.
In Fig. 2 and Fig. 3 are presented examples of fatigue bending tensile stress-deflection
curves respectively related to a specimen which has failed before 2 millions cycles and to
a specimen which has not failed.
In the Fig. 4 are presented examples of deflection-cycles number curves respectively
related to the case where the specimen is broken before 2 millions cycles and the case
where the 2 millions cycles were reached. We find the usual shape of the fatigue curves
with three different phases:
- a first phase, corresponding with a starting micro-cracking of the matrix. The deflection
evolution is fast.
- a second phase marked with a slowing down of the deflection evolution.
- a third phase which marks the resumption of the damage and leads very fast to the ruin
of the structure. This last one is of course absent for specimen weakly damaged.

On the Fig. 5 is represented the fatigue cycles number-load rate (R) diagram. From this
figure one makes the following remarks :

- results are relatively scattered, that is usual for fatigue tests ;

- below a loading ratio R = 0.8 specimens in CEMTECmultiscale ® do not fail by fatigue
before 2 millions cycles.

The stress scattering observed on the Fig. 5 indicates that the applied load ratio is not
good parameter to analyze the fatigue failure probability related to composite. The
applied stress/MOR ratio related to each specimen is surely a best parameter to evaluate
this fatigue failure probability of a specimen of CEMTECmultiscale <E. Not being able to
determined this ratio, the specimen MOR not being known, the initial static damage
(quantitatively represented by the initial static strain) generated during the first static load
could be a good parameter to evaluate this fatigue failure probability.

To show this dependence, we consider in a first step S, the derivative of Deflection

evolution-cycles number curve. The S-initial strain diagram is presented in Fig.6 This

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

96 Parant and Rossi

figure shows a good correlation between the stationary fatigue damage phase and the
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

initial static damage.

Considering this positive result, the fatigue failure cycles number-initial strain diagram
is, in a second step, drawn (Fig. 7). If we consider the Fig. 7, we can make the following
comments:

- There is a critical initial strain threshold below which specimens do not break before 2
millions cycles, while below this threshold the rupture becomes inevitable. This
threshold value is between I.27 X I 0" 3 and 1.44 X I 0"3•

- Beyond the threshold, there is a linear relation between the fatigue failure cycles
number and the initial static strain.

Specimens having reached the 2 millions cycles were then reloaded in quasi-static test
until rupture. The Fig. 8 presents the five static reloading curves, as wel1 as the average
curve.

Average static curves, before and after 2 millions fatigue cycles are represented in the
Fig. 9. For more legibility, we incorporate the min. and max. reloading static curves. We
observe that the static bending behavior after 2 millions fatigue cycles is better than those
related to the specimens not loaded in fatigue. The gain is about 6.5 %. The deflection at
the strength peak is approximately the same in the two curves.

Finally, if we consider the characteristic stress related to a strain equal to 1.27 x 10"3,
which is the lower value of the critical initial strain evocated above, we obtain a
characteristic stress equal to 30 MPa, that corresponds to a loading ratio R of 0.65.

CONCLUSION

A new ultra-high performance cement composite, the CEMTECmuttiscate ®' was tested under
static bending, stactic compression and under asymmetric fatigue bending. From this
experimental study, the following comments can be made:

I. The average strength and ultimate strain in compression are respectively equal to 220
MPa, and 4.5 10·3 •
2 The characteristic strength and ultimate strain in compression are respectively equal
to 205 MPa, and 4 I 0" 3 .
3. The Young modulus is equal to 55 GPa and the Poisson coefficient is equal to 0.21.
4. The average modulus of rupture (MOR) is equal to 61.5 MPa;
5. The average strain related to the average MORis equal to 9.2 10"3•
6. The stationary damage fatigue evolution of a specimen is dependent of the initial
static damage of this specimen.
7. A critical initial static strain threshold exists. Before this threshold a specimen in
CEMTECmultiscale ® does not fail during a bending fatigue loading and beyond this

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 97

threshold the failure fatigue cycles number linearly depends of the initial static strain.
3 3
The strain threshold determined in this study is between 1.24 x 10- and 1.44 x 10- •
8. Below a loading ratio R = 0.65, failure during bending fatigue test never appears with
a specimen ofCEMTECmutriscate ®.
9. A gain of 6.5 % is observed between the bending static behavior of the specimens
being previously loaded in fatigue and those not being loaded in fatigue.

ACKNOWLEDGEMENTS

The authors thank the team ofS. RICORDEL, namely J.D. SIMITAMBE, F. GUIRADO.
CARR lA T for the important work supplied during the mixing operations and preparation
of specimens.

REFERENCES

Bache, H.H., 'Compact Reinforced Composite. Basic principles', Aalborg Portland,

Cemcnt-og Betonlaboratoriet, CBL report 41 (1987).

Richard. P., Cheyrezy, M., 'Les Betons de Poudres Reactives', Annales de l'ITBTP 532
(1995) 85-102 (in french)

Rossi P., "High performance multimodal fiber reinforced cement composite

(HPMFRCC): the LCPC experience," ACI Materials Journal, 1997, vol. 94, n°6, pp.
478-483.

Lankard, D.R, Newell, J.K 'Preparation of highly reinforced steel fibre reinforced
concrete composites', Fiber reinforced concrete, SP-81, (American Concrete Institute,
Detroit, 1984) 286-306.

Li, V.I., 'Engineered Cementitious Composites - Tailored composites through

micromechanical modeling', Fiber Reinforced Concrete: Present and the ji1tur,
Montreal, Canadian Society for Civil Engineering, (1998) 64-97.

Rossi, P., Acker, P., and Malier, Y., "Effect of steel fibers at two stages: the material and
the structure," Materials and Structures.l987, vol. 20, pp. 436-439.

Table 1- Mix des1gn. of the CEMTECmutriscute

Cement 1050.1
Silica fume 268.1
Sand 514.3
Water 180.3
Suoemlasticizer 44
Steel fibers 858
Water/cement= 0.201
Water/ binder= 0.16
Air entrained = 20 litres
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

98 Parant and Rossi

~~------------~----------------

:f 70 i Static Average curve

~ =JI
~ 40 .
f!
J!l 30 '
CD
.5
I

l \
Characteristic curve
Min.

~.,
Ill

0~-----~----~----~----~----~----~
0 2 4 6 8 10 12
strain [x 1 o;
Figure 1- Static Bending tensile stress-Strain curves

50 First cycles
cycles (5000)
Oi'
D.
40
.
i!.

..
t
.!!
30

·;;;
c 20
s
01
c
:g
.
c
Ill
10

0
0 2 3
deflection [mm)

Figure 2---Cyclic Stress-Deflection curves until rupture

-----------------~------------

first cycles cycles cycles

Oi'40 cycles (5 000) (25 000) (100 000)
D.
i!.
:230
;
..
"'
:;;;20
c
.!
m
c

..
;:; 10
c
,g

0
0,0 0,4 0,8 1,2 1,6
deflection [mm]

Figure 3- Cyclic Stress -Deflection curves with stagnation

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 99

2,8

1.4 x 10.,; (rrrn I cycle)

2,4

e
.sc 2.0
4.1 x 10"8 (mm I cycle)
0
=... ~

Q
..
.!
1,6

1,2
O.OE+O 5,0E+5 1,0E+6 1,5E+6 2,0E+6
Cycles number

Figure 4. Deflection evolution-Cycles number curves

2,4E+6 T'------------------,
...
E 2,0E+6 X X X
~

..
:;_ 1,6E+6

~
~ 1,2E+6

-...
u

~ 8,0E+5
.a
~ 4.0E+5 X
z
X .'><': ./ -~- ~/V
O,OE+O f-'-~-t-'-......._~..l.......'-~-'........_.....;s.,~-'--'--'--+*""'~~.....;;~""*-'-'-1
0.70 0,75 0,80 0,85 0,90 0,95 1,00 1,05 1,10
stress ratio (Fatigue Stress..Characteristic Stress)

Figure 5- Fatigue cycles number-Stress ratio diagram

1E-8

1E-7
LOQ1o (S) =-10.42 X l':j + 2,85
1E-6
Ill
....
0
1E-5 +
iii 1E-4

1E-3
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

1E-2
0,5 1,0 1,5 2,0 2,5 3,0
3
Initial Strain E I [X 10" I

Figure 6 - Deflection evolution Slope-Initial strain diagram

Copyright American Concrete Institute

100 Parant and Rossi

1E+7

...
!!5,1E+5
1E+6 +tl
l
'
+
Log1o ( Nc) =·1,47 XEi + 7,38

~ i
..
>-
~ 1E+3
1E+4
l!
•
~ 1E+2
I
!
z= 1.27 ! 1.44
1E+1

1E+O
1.0 1,5 2,0 2,5 3.0
Initial Strain£( [x 10i

Figure 7- Initial strain-number of cycles (at rupture) diagram

0 2 3 4
Deflection [mm]

Figure 8 - Static behaviors after 2 x 106 cycles

80T-------------------------------~

:70
:I
.
';;' 60

..
.r:
:!.,
50
Befol& fatigue

s
. 40
30
badi~ (average)

~"'
..
Ill 10
20

0~--~--~----~---r--~----r---~--~
0 2 3 4
deflection [mm)
Figure 9 - Static behaviors before and after fatigue loading

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

SP-224-08

Structural Evaluation of Cement Skin

Sandwich Building System

by Y. Shao, E. Blain-Cosgrove, and B. Robinson

Sl·nopsis: The balance between sustainability and affordability is hard to achieve when

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
considering choices of building envelopes. A simple and easy-to-construct stressed skin
structural sandwich system that is both affordable and sustainable is evaluated in this
paper. The system is composed of an expanded polystyrene (EPS) panel core, wrapped in
polymer mesh and covered with a thin cement skin on both sides. This system design
leads to a highly energy efficient building envelope system. A full-scale sandwich wall
was constructed and tested to examine the possibility of its use as a load bearing wall in
one story residential house without traditional timber frames. Based on the requirements
imposed by the National Building Code (NBC), the test results from this experimental
program were found to be promising. The wall carried a gravity load, a wind load and
seismic in-plane shear load at least 4 times as high as the NBC design load with negligible
lateral displacement and no visible cracking. At buckling failure, the load-carrying capacity
of the wall exceeded 10 times the design load. The EPS-core stressed-cement skin
sandwich building system thus provides a good example of the use of thin cementitious
products in load bearing exterior wall structural applications.

Keywords: expanded polystyrene; frameless housing; full scale wall tests;

gravity load; sandwich building system; seismic resistance; thin cement
products; wind resistance

101
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=University of Texas Revised Sub Account/5620001114, User=erur, ert
No reproduction or networking permitted without license from IHS Not for Resale, 01/26/2015 02:27:05 MST
Daneshlink.com
daneshlink.com

102 Shao et al.

ACI member Yixin Shao is an associate professor of Department of Civil Engineering
and Applied Mechanics at McGill University, Montreal, Canada. His research interests
include fiber reinforced cementitious materials, fiber reinforced polymeric materials,
processing and building product development. He is a member of ACI 549, 544 and 236.

Emmanuel Blain-Cosgrove is the President of Ecohabitation.com of Montreal, Canada.

He has been extensively involved with sustainable development of environmentally
friendly housing for remote areas and developing countries.

Brad Robinson is the founder of Intematural Canada of Quebec, Canada. He has been
working on the energy efficient housing design and unitization of waste materials for 20
years. His innovation includes passive solar system, straw bale housing and bio-blocks.

INTRODUCTION

The sandwich structures are extensively used in airplane and automobile industry for
lightweight, high strength and stiffness, and economic design [1-3]. In construction
industry, this concept has been slowly but increasingly accepted. Sandwich roofing
panels made of expanded polystyrene (EPS) core bonded with plywood or flake board on
two sides are commercially available on the market. The straw bale houses and bio-block
constructions were all based on the cement skin sandwich building system [4]. The
growing use of straw bale for load-bearing building envelopes has made a significant
impact in the field of sustainable housing. Load-bearing straw walls have been praised for
being affordable, having high compressive strength and thermal properties, having
minimum environmental impact and reducing the use of valuable lumber. However, the
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

process of building homes with straws is often painstakingly long and labor intensive,
requiring considerable work to compress the straw materials for energy efficient bio-
block cores. One of the solutions is to use the expanded polystyrene panels ("bead
board") as a core material in this cement skinned sandwich system.

The EPS-core sandwich wall is intended to deal with many of the challenges of
sustainable building: load-carrying capacity, affordability, energy efficiency, minimal
environmental impact materials, and occupant health. The cement skin can be applied
manually on site or prefabricated in factory, the former being suitable for remote areas,
while the latter for rapid construction of urban houses and commercial buildings.
Considerable savings are incurred when comparing the system to conventional wood
frame construction. The wall system is economical in the long term as well in reducing
heating and cooling costs by providing a tight and uniform insulation.

This paper reports a study on the load-carrying capacity of the cement skin - EPS core
sandwich building system. A full-scale wall test was conducted to examine the structural
response to the gravity, wind and seismic loads and make a comparison with the design
loads specified by the National Building Codes.

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 103

MATERIALS

Mortar mix with sand to cement ratio of 2.5 and water to cement ratio of 0.6 was selected
as cement skin. The 7-day compressive strength was 19.4±0.14 and 28-day compressive
strength 20.4±1.03. With normal air curing, there was no significant difference between
7-day and 28-day properties. Therefore the full-scale wall tests were carried out 14 days
after thin cement applied and cured in the air.

EPS bead board was the insulation of choice for the project. In contrast to extruded
polystyrene, where HCFCs are used as the blowing agent, "bead board" EPS foam panels
are made with water vapor-blown polystyrene. Furthermore, the product used for this
research was made from over 80% pre- and post-consumer recycled polystyrene, that is
easily recyclable at the end of a building's life cycle. Finally, the materials used in the
system are relatively benign with regard to off-gassing, during and post-construction,
making it a good choice for a "healthy" building envelope. It costs almost half the price
of extruded foam. The board as received was I 0 I mm (4") thick. Three pieces were
bonded together to make up the core of 305 mm ( 12"). The compressive tests were
performed to determine the maximum load capacity of the core. It was about 12.5 kN/m
(852 lblft). The thermal resistance of the EPS bead board is about R = 3 h.ft2.F/Btu/in.
For 305 mm ( 12") thick EPS core, the thermal resistance of the wall is R = 36. This is a
high performance wall in a very cold climate. The building codes require that R == 20 be
the standard in Canada.

Polypropylene-based plastic mesh was used as a wrapping reinforcement for EPS core, as
a connection between core and skin, as well as an internal reinforcement for cement skin.
The mesh size was 2.3 strands per 25.4 mm (I in) and same in two mesh directions. The
breaking load was 2816 N/m ( 193 lb/ft).

EXPERIMENTAL PROGRAM

An EPS sandwich wall of 2.75 m (9') tall, 1.22 m (4') wide and 0.35 m (14") thick,
including the 25 mm (I") cement skin on each side of the wall, was constructed for full-
scale wall test. The purpose was to determine the capacity ofthe wall to resist the vertical
gravity load, the horizontal wind load and the cyclic shear load.

In order to obtain more information from the full-scale wall tests, the sandwich wall was
constructed first with a closed edge box-section (Fig. I a) to simulate the end condition of
a house and then with a cut open edge (Fig. 1b) to represent a mid-unit of the wall.

For the closed edge section, tests were conducted to examine the performance of the wall
when the applied load is first gradually increased to the design load level, and then
continued to four times the design load. The gravity load tests were performed first,
followed by wind load tests and in plane shear tests. Each test underwent three load-
unload cycles and visual inspection was carried out to detect the possible cracks.

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

104 Shao et al.

For the open edge section, similar procedure was followed to repeat the tests with weaker
cross section. The gravity load tests were performed again to the design load and then to
four times design load as well with three cycles of load and unload for each. The wind
pressure was applied afterwards to the design load and gradually increased until the first
transverse cracking. The cracked wall was finally tested to buckling for maximum load
bearing capacity at failure.

According to NBC (1995), a single story residential house of 6.7 m x 13.4 m (22 ft x
44ft) in Montreal area should be designed to carry the following loads:
Gravity load on 6.7 m (22ft) wall= 38.5 kN/m (120 lb/ft 2 x 22ft= 2640 lb/ft)
Gravity load on 13.4 m (44ft) wall= 19.3 kN/m (120 lb/ft 2 x 11ft= 1320 lb/ft)
Wind load= 1.2 kPa (24 pst)
Seismic load = 2.36 kN/m ( 162 Jb/ft)

On a 1.22 m (4ft) wide wall, the design loads are:

Gravity load on 6.7 m (22') wall= 48 kN
Gravity load on 13.4 m (44 ') wall = 24 kN
Wind load = 1.2 kPa (24 pst)
Seismic load= 2.9 kN (648Jb)

CONSTRUCTION OF SANDWICH WALL

Full-scale sandwich wall was constructed on the strong floor directly under the 104 kN
(2.2x I 06 1b) MTS machine in McGill University's Structural Lab. concrete footing (3m x
1m x 0.15m) was cast first with anchor bolts in position. Steel meshes were embedded in
the concrete footing, extending 0.3 m above the footing on two sides (Fig. 2). Four days
after casting the concrete footing, the EPS core was placed on the top of the footing and
guided by a wood frame to control the thickness of the cement skin. The core was
wrapped by plastic mesh and connected to the footing by the 0.3 m tall steel mesh. The
cement skin was applied manually to simulate the site condition in remote areas. Two
additional layers of plastic mesh were added; one close to the EPS core and the other near
the surface. The finished wall is shown in Fig. 3 with top end braced by timbers, the
bottom end bolted to the strong floor and front surface painted white to monitor the
cracking. A 25mm thick plywood plate was placed on the top of the wall to evenly
distribute the gravity load.

An air bag was built using a polypropylene plastic sheet in a wooden frame braced by
timbers and supported by steel columns of the MTS machine. The air pressure in the bag
was measured by the differences of water height. The schematic of the set-up for gravity
load tests and wind load tests is shown in Fig. 2. Two LVDTs were used to record the
displacements of the mid-point of the wall. LVDT 18 measured the front surface
displacement and LVDT 17 the back surface (air bag side) displacement. The latter was
accomplished through a steel rod embedded inside the wall and glued to the back surface
by epoxy.

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 105

The set-up for in-plane shear tests is shown in Fig. 4. Two hydraulic jacks, two load cells
and two L VDTs were installed to carry out the in-plane cyclic loading. The top end of the
wall was clamped by a steel frame to allow the push from two sides by the jacks. Two
rollers were added on the top under the universal joint to provide a mechanism for
movement while the wall was simultaneously subjected to a gravity load. When jack 1
was pushing, LVDT 2 recorded the displacement. While jack 2 was pushing, LVDT 1
collected signals. The cycle was done at least three times for each design load.

RESULTS

Full scale tests of wall with closed edge

Fig. 5 shows the response of the wall to the cyclically added gravity load up to 200 kN (4
times design load); no wind pressure (lateral load) was applied. Thick lines corresponded
to loading while the thin lines to unloading. Very small lateral displacement was
detected, indicating the rigidity of the wall was maintained during the load-unload cycles
to four times the design capacity.

The response of the wall to the cyclically added wind pressure up to 5 kPa (4 times
design load) is demonstrated in Fig. 6. A 48 kN dead load was kept on the top. Both
LVDT 17 and 18 were used to detect displacements. It seemed that the back side (wind
side) displacement detected by LVDT 17 was larger than the front side (L VDT 18),
suggesting that the two cement skins did not move by a same amount. The residual
displacement after the first cycle was almost negligible. No crack was noticed.

Figs. 7 displays the wall response to the cyclic in-plane shear load up to three complete
cycles and at a load of 11.6 kN (4 times design load). The tests represent the expected
loads due to a seismic event on the house. The roof dead load of 5.1 kN was applied as
constant gravity load. No damage to the wall was observed.

Full scale tests of wall with open edge

To examine the structural behavior of the wall only with two face skin layers, the 6 mm
thick cement on two sides was cut open. Tests were repeated on the wall with open edge
to simulate the mid-wall unit in a building envelope. Figs. 5 - 7 have showed that there is
no significant degradation in strength and stiffness of the wall with closed edge after
cyclically loaded to four times design loads.

Fig. 8 exhibits the results of cyclic gravity load tests after three cycles at 200 kN. The
maximum lateral displacement monitored by two LVDTs was 0.1 mm, indicating the
wall with open section was still stable and rigid, and could be used for failure analysis. 1t
was interesting to compare Fig. 8 with Fig. 5. The two skins in open edged wall tended to
move away from each other (Fig. 8), while the two skins in closed edged wall always
moved in the same direction. The two lateral displacements (Figs. 5 and 8) were of the
same order of magnitudes.

Fig. 9 demonstrates the cyclic wind load test results up to the wind pressure of 4.5 kPa
and 5.5 kPa. At 4.5 kPa, the wall with open edge exhibited similar response as the wall

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

106 Shao et al.

with closed edge. The first transverse crack was initiated at 5.5 kPa on the front tensile
surface with significantly irreversible lateral displacements. The gap between the air bag
and the wall was apparent due to the deflection of the wall. With open edge, the wall
underwent approximately 10 times more deflections than that observed with the closed
end. The differences between the two L VDTs were observed to decrease when larger
lateral displacements occurred, suggesting that the bond between the cement skins and
the EPS core by the plastic mesh was maintained.

Compressive tests were conducted to examine the vertical load carrying capacity of a
cracked wall. The wall was loaded up to 300 kN, unloaded to zero and then loaded again
to failure (Fig. 10). Significant lateral displacement was observed. In failure tests, the air
bag was removed and LVDT 17 was placed on the back side. The two LVDTs exhibited
almost identical values. The typical failure mode is buckling. A second transverse crack
was also observed during the buckling yielding. The test was stopped when the maximum
mid-height displacement reached 38 mm. The load capacity of the cracked wall was 406
kN/m.

CONCLUSIONS AND RECOMMENDATIONS

A full-scale EPS core-cement skin sandwich wall was constructed and tested to examine
the possibility of being used as load bearing wall in one story residential house. The test
results were promising.

For both closed edge and open edge cross sections, the wall could carry a gravity load of
164 kN/m with barely any lateral displacement. This is 4 times higher than the required
NBC design load (38 kN/m). At buckling failure, the capacity reached at least 406 kN/m.

In wind pressure tests, the closed cross section had shown resistance to lateral
displacement almost 10 times higher than the open section. Both sections could carry at
least a wind pressure of 5 kPa ( 100 psf); about 4 times the design load ( 1.2 kPa ).

The seismic resistance of the wall was investigated by in-plane cyclic shear tests. The
maximum load of 11.6 kN was applied to represent 4 times the design load (2.9 kN). The
corresponding hysteretic permanent deformation was approximately 1 mm in a 2.7 m tall
wall.

The cement skin EPS core wall system demonstrated sufficient strength to serve as load
bearing exterior wall for one-story residential housing without traditional wood frames.
For a frameless house using this sandwich system, the joint design and construction
between the walls and the wall to roofing play a critical role and need to be investigated.

The high load carrying capacity of the wall system is attributed to the fully developed
strength in thin cement skin stabilized by the core. The maximum average load was about
13 kN/m for EPS core and about 1550 kN/m for sandwich block with two 25 mm thick
skins. The contribution of the EPS core to the structural capacity appears to be negligible.
Therefore, any materials that can hold the skins may be used as a core in the proposed
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 107

system. This represents a wide range of opportunities for designers to select core
materials to construct environmentally friendly, thermally efficient, structurally strong
and economically feasible residential houses and commercial buildings.

REFERENCES

[I] Allen, H. G. (1969), "Analysis and design of structural sandwich panels", Pergamon
Press, London, UK
[2] Hockman, L. E. (1973), "Sandwich construction and design", Analysis and design of
flight vehicle structure, Bruhn, E.F. ed., Jacobs & Associates, Indianapolis, Indian.
[3] Platema, F. J. (1973), "Sandwich construction", John Wiley & Sons, New York, N.Y.
[4] Robinson, B. (1996), "Proof of concept: development and testing of the biocrete
house construction system", Intematural Canada. Canada Mortgage and Housing
Corporation, Ottawa, Canada.

ACKNOWLEDGMENT

The authors gratefully acknowledgment the financial support from Canada Mortgage and
Housing Corporation (CMHC), and the team in the External Research Program, who
made the project possible.

0.35Sm (14") EPScore

6mm

1.2 m (4')
I

(a) Cross section ofEPS sandwich wall with closed edge

25mm

•.,.
0.355m (14") EPScorc
-too !'f- 6mm

1.2m(4')
J
(b) Cross section ofEPS sandwich wall with open edge

Fig. I: Cross section of EPS sandwich wall

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

108 Shao et al.

Steel block

2.75m (9')

1.37m (45')

0.15m{6"''
Concrete footing

Fig. 2: Side view of sandwich waJI (waJI thickness= 0.355m)

Fig. 3: Finished waJI

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 109

LVOTI l\'OT2

!.23m

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
Fig. 4: Front view of sandwich wall

~r------------------------------------------,
:~tvtf!t1 .. ~1Q~~fbtti-W~Ihit$~i:l:4d
z«;~~-..~~te~t) ~~biN*lMt)'l~-E~Ggf:t
l.VOT18

... e.l c4 c.t~ t».6 ~7 0.$ 0.9

--~---
Fig. 5: Cyclic gravity load tests (closed edge)

WiM ........... up to $11P0(1Np$f)Mthlt wd•JthdaMO._.,,

TopkMd•UkM

_...... LV0ft7
• L\.'DJta

,.
¢2 c• oE.

---
oe

Fig. 6: Cyclic wind load tests (closed edge)

,2 ~4

Copyright American Concrete Institute

110 Shao et al.

'¥'<' ........... _.,...

V»tU.N~~t
..........10··f·. ~~~lt(I~2Pt-H3P3f).
~ic).adontopof1.2l~n~')'MI .. S..1kN
(11.U:b).

Fig. 7: Cyclic in-plane shear tests (Closed edge)

I
~fQQiDnumt to~ tbl!

-""""'""'........ -
g:rM,Iold

!
I
·0.15 O:Hi

Fig. 8: Cyclic gravity load tests (open edge)

~ .... lvunr:
·•··l\'01'1•:
["!i:!f~:~...~ef_J

w
WJnd ..,._.. . •u~
Ot\tN Wlllwttt\ope....,..,
,.., ...... kM

•• IS

Fig. 9: Cyclic wind load tests up to failure (open edge)

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 111

·SOO

......
.
!!
j

~
...,
i
~
u
...
M d ~ n w » ~ o
MfiS.PiomtLatat41~n.mM

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
Fig. 10: Failure test of cracked wall up to buckling (open edge)

Copyright American Concrete Institute

112 Shao et al.

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

SP-224-09

Properties of Short Fiber Reinforced Cement

Paste for Concrete Tubes Produced by
Centrifugation Method

by D. Hesselbarth and J. Kaufmann

Synopsis: Concrete tubes are usually produced by a centrifugation method using steel
bar reinforcements. The reinforcement of concrete with steel bars is expensive, susceptible
to corrosion and leads to rather thick and heavy structural elements. The application of
short fiber reinforced cement (FRC) or mortar is a suitable alternative. The paper presents
the development and evaluation of a suitable FRC for this particular application. First, the
cement matrix was optimized for use in a conventional casting forming process. A mixture
of ultra-fine cement and ordinary Portland cement improves the rheological properties of
the fresh mixture and results in a very dense cement matrix with excellent mechanical
properties. This optimized cement matrix was then reinforced with different kinds of
carbon and polymeric fibers such as PYA and PP. Hereby, the carbon fibers primarily
increase the llexural and tensile strength of the material, whereas the polymer fibers tend
to improve the ductility of the cement matrix. Furthermore, the influence of water-reducing
agents, of different constituents (microsilica, filler, sand), and the mixing process on the
mechanical properties were studied. The mechanical properties were found to depend also
on the curing conditions of the hydrated samples. The microstructure and the fiber-matrix
interface were investigated by ESEM (Environmental Scanning electron microscope). In a
further test series, the mixtures were optimized with regard to the flow properties needed
for the centrifugation process. The mechanical properties and the microstructure were
investigated. As a result, this work shows the possibility to apply the FRC for industrial
production of centrifuged tubes.

Keywords: carbon fibers; centrifugation; fiber reinforced cement; flexural

and compressive strength; polymeric fibers

113
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

114 Hesselbarth and Kaufmann

Daniela Hesselbarth received her Master degree in civil engineering from the Slovakian
University of Technology, Bratislava, Slovakia, in 1992, and the Ph.D. with a thesis
about alumina ceramics from the Swiss Federal Institute of Technology, Zurich,
Switzerland, in 1999. She is now with the Swiss Federal Institute for Materials Testing
and Research (EMPA), working on fiber reinforced composites.

Josef Kaufmann received his M.Sc. in physics and the Ph.D. with a thesis about frost
damage mechanisms in concrete from the Swiss Federal Institute of Technology Zurich
and Lausanne, Switzerland, in 1989 and 1999, respectively. He is actually a senior
researcher at Swiss Federal Institute for Materials Testing and Research (EMPA).

INTRODUCTION

Cementitious materials are known to be very brittle in tension, with small strength and
strain capacities. Steel reinforcement is usually necessary in order to use cementitious
materials and concrete as a construction material. The reinforcement of concrete load
bearing elements and secondary elements with steel bars is time consuming and
expensive, susceptible to corrosion and leads to relatively thick and heavy structural
elements. The substitution of the steel bars by high performance short fibers may allow
producing thinner, lighter and cheaper elements. Fiber reinforced cementitious (FRC)
materials show very high strength in flexure and tension with high ductility.

A prerequisite for the development of FRC is an appropriate, optimized cement matrix.

The matrix should combine properties such as low porosity, high density, and excellent
mechanical properties. The matrix can be improved by combining two or more
components with different size distributions [1]. The addition of micro-fine cement
(d 50 < 3J.1m) to the ordinary Portland cement improves the rheological properties of the
fresh mixture and leads to a dense cement matrix with high strength.

Short, high performance fibers are known to increase the flexural and tensile strength as
well as the ductility of the cement matrix. A volume fraction of Jess than 5% of carbon
fibers results in a strong reinforcing effect due to the exceptional mechanical properties
(Young's modulus, tensile strength) of the fibers. High performance PYA
(polyvinylalcohol) fibers are used to produce composites similar to asbestos reinforced
materials for asbestos replacement [2]. PYA composites with highest toughness and
strength can be produced, at the cost of somewhat lower E-modulus than that of the
asbestos cement composites [3]. The application of PP (polypropylene) fibers with their
modest mechanical properties is predominantly found as a secondary reinforcement with
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 115

small volume fraction (<1 vol.-%) for crack control. When used as primary reinforcement
for strengthening and toughening, a high volume fraction of the PP fibers and special
forming techniques are required [2].

From the processing point of view, the workability of the fresh cement mix containing
fibers is very important. Fibers, in particular carbon fibers, tend to stiffen the mix. The
problem of obtaining good fiber dispersion and consistency gets more complicated for
very low water to binder ratios (w/b < 0.25) and for the centrifugation process. One
solution consists in replacing a part of the cement by silica fume [4] and/or adding latex
dispersion [5,6]. Both additives improve the dispersion of carbon fibers in the fresh
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

mixture and reinforce the bonds between fibers and matrix.

For the centrifugation process, the fiber cement mixture is required to be thixotropic,
which means to flow during the spinning process and to get stiff after the centrifugation.
The fibers should move well during the centrifugation, in order to reach a homogeneous
distribution in the matrix. Care has to be taken that no segregation of any mixture
component occurs during centrifugation.

MATERIALS

A binder system contained Ordinary Portland Cement (CEM I 42.5 N, d50 = 24Jlm) and
. micro-fine cement (Portland cement blended with blastfumace slag with d50 = 2.8Jlm). A
superplasticizer based on polycarboxylate was used as a water reducing agent.

The fiber dispersant was a polymer dispersion based on styrenbutadien. Three types of
PAN (Polyacrylnitril) based carbon fibers with similar mechanical properties but
different surface properties were evaluated (CF no. 1-3) and pitch based carbon fiber (CF
no. 4). Furthermore, PYA fibers and PP fibers were investigated. The fiber properties are
listed in Tab. I. For comparison, silica fume instead of micro-fine cement was used in a
number of mixtures.

EXPERIMENTAL PROCEDURE

The composition of the mixtures was as follows: The binder was composed of 80 wt% of
CEM I 42.5 N and 20 wt% of micro-fine cement. Three different w/b ratios 0.18, 0.20
and 0.22 were used. The amount of superplasticizer ( 1.5, 2, 2.5, 3 wt% of binder) was
combined with each w/b ratio. They were mixed according to the European standard EN
196-3 for 3 minutes in a Hobart machine. First, the fibers were evaluated using a

Copyright American Concrete Institute

116 Hesselbarth and Kaufmann

conventional casting procedure. Samples (25x25x100mm 3) were cast and stored for 24
hours at 20°C, 95% RH. After demolding, the samples were cured in water at 20°C. The
flexural strength was estimated in three point bending test (span 80mm) on series of 6
specimens. The compressive strength was measured on cubic samples (size 25mm) after
7, 28, and 90 days on series of 6 specimens.

Centrifugation or spinning at rather high speed is an integral part of the manufacturing

process of reinforced or pre-tensioned concrete poles, piles and pipe_s. The compaction
due to the centrifugation can give denser and stronger products as well as a better surface
finish. If the process parameters are carefully chosen, segregation effects can be
minimized. For fiber reinforced materials, the spinning may change fiber distribution and
orientation. The latter effects can even be used to give advantageous material properties
[7]. The mixtures were first mixed in a Hobart machine. For the spinning process, they
were then transferred into a conical steel form (length 350mm, diameter varying from
180mm to 176.5mm). Some parameters like w/b ratio, spinning speed, addition of silica
fume and different amount of latex were varied. The mixtures have to be optimized in
regard to their special consistency (thixotropy), which is necessary for the spinning. The
mixing proportions are given in Tab. 2. The centrifugation at 570 or 669 rpm took 10
minutes.

The samples were demolded after 24 hours and afterwards cured in water. After five
days, samples of 15x 15x80 mm 3 were cut and after 7 and 28 days, the mechanical
properties on series of 15 specimens were measured.

RESULTS

Conventional Forming - Casting

The effect of the addition of micro-fine cement and superplasticizer on the mechanical
properties of the cement matrix was investigated. The addition of micro-fine cement to
the Ordinary Portland Cement improves the rheology of the fresh mix and the mechanical
properties of the matrix, in particular the flexural and compressive strength (Fig. 1). The
porosity of the matrix decreases, whereas its density and durability increase.

The influences of the superplasticizer and w/b ratio on the flexural strength are shown in
Fig. 2. The increase ofw/b ratio decreases the flexural strength, whereas a higher amount
of the superp]asticizer results in a decrease of the flexural strength. This may be due to an
increase of the particle packing density of the cementitious matrix caused by the
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 117

superplasticizer polymer. A dense structure is more sensitive to shrinkage induced micro-
cracking, resulting in the observed decrease of the flexural strength. The contrary effect is
observed on the compressive strength, which increases with higher content of
superplasticizer (see Fig. 4).

Some mixtures (depending on the w!b ratio) show a slight decrease of the flexural
strength with age (Fig. 3.) The continuing hydration of non-hydrated cement grains
occurs by the diffusion of water which may develop internal stresses. As a result,
cracking may occur and strength and durability will decrease. Curing under water also
seems to induce internal stresses. Water transport is limited through such a dense matrix,
and inhomogeneous hydration may occur. This effect can be avoided by the application
of an adequate curing regime [8]. As shown in Fig. 4, the compressive strength increases
with age.

Properties of the fiber composites

Different high performance carbon fibers were added to the binder matrix (Ordinary
Portland Cement blended with micro-fine cement) and the mechanical properties of the
composites were investigated. The w!b ratio was 0.20, the superplasticizer amount 3wt%
to binder, amount of fibers 2vol%. A latex dispersion was added at an amount of 2wt% of
binder.

As shown in Fig. 5, the flexural strength of the cement matrix can be increased by adding
carbon fibers. The amount of the improvement is known to depend on a number of
parameters, such as fiber length, orientation, fiber-matrix bond strength, as well as on the
mechanical properties of the fibers (2]. It was found that the flexural strength of the
composites show·n in Fig. 5 correlates with strength and Young modulus of the fibers:
PAN based carbon fiber CF no. 3 exhibited the highest strength and modulus (see fiber
properties in Tab. 1.), whereas the pitch-based carbon fiber CF no. 4 showed rather poor
mechanical properties. The somewhat better performance of CF no. 2 after 7 days is not
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

statistical significant.

A critical point seems to be the fiber-matrix bond. The sizing of the fibers affects to a
large extent the flexural strength of the composite. On the other hand, the fiber surface
treatment changes the mixing properties of the slurry. For example, the PAN based
carbon fibers no. I have a glycerin sizing, making mixing easy. However, the fracture
ESEM picture (Fig. 6) shows that there are apparently no large areas of contact between
carbon fibers no. I and the cement matrix. Contrary to this, the PAN based carbon fibers

Copyright American Concrete Institute

118 Hesselbarth and Kaufmann

no. 2 have a thermoset sizing. The ESEM picture (Fig. 7) indicates that there are indeed
contact areas linking the fiber to the cement matrix.

The addition of different amounts of latex dispersion and their influence on the flexural
strength of the composite were investigated in a separate series. At the constant mix
proportion with w/b ratio 0.2, superplasticizer 3wt% of binder, carbon fibers (CF no. 2)
2vol. %, the amount of latex was varied from 0, I, 2, 3, 5, 15 and 20wt% based on binder.

The addition of latex is known to improve the fiber dispersion in the fresh cement mix
[9]. Poor fiber dispersion was observed when latex was totally absent. In addition, the
latex produces a major improvement in the bond strength [5]. In our case (Fig. 8), a
high amount of latex (more than about I Owt% to binder) modifies the cement matrix,
resulting in high flexural strength until a plateau is reached. A small amount of latex
(1 ... 5wt% to binder), however, mainly affects the fiber dispersion. The highest flexural
strength is obtained with 2wt% to binder. A somewhat higher latex content (3 ... 5wt %)
may delay the hydration by covering the cement grains, thus reducing the strength.

The dependence of fiber efficiency in dependence on fiber type was investigated m

special series. The carbon fibers no. 2, PVA fibers, and PP fibers of length 6 mm,
respectively, were used in these experiments. Good workability requires a suitably
chosen amount of fibers in the mix. Tab. 1 shows the fiber properties. The compositions
of the mixtures are given in Tab. 3. Figs. 9 and 10 show the results of these
investigations. The efficiency of the fiber reinforcement (enhancement in strength and
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
enhancement in toughness of the composite, compared with brittle matrix) depends on
fiber type, and the mechanical properties of fibers themselves. Furthermore, the fiber
length, fiber orientation and fiber-matrix bond play important roles on the efficiency of
fiber reinforcement. The carbon fibers with their excellent mechanical properties are very
effective on the strength increase of the composite, where the flexural strength reaches
more than 30 MPa. The flexural strength of composites containing PVA or PP fibers
(note the high volume contents of these fibres) reaches about 19 MPa. A load deflection
curve (Fig. 10) shows the effect of the fibers on the toughness of the composites and its
crack control potential. Of particular importance in the load deflection curve is the post
cracking zone, which represents the strain capacity and toughness of the composite, and
thus provides an indication of the quality of the material from the point of view of crack
control. The three fiber types (CF; PYA and PP fibers) show profound differences in the
post-cracking zone.

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 119

The carbon fibers exhibited the first crack at very high flexural strength, but poor strain
capacity in the post cracking state. The PP and PVA fiber composites did not differ much
in their (moderate) flexural strength; they showed a difference in the post cracking
behavior. They have a good strain capacity and toughness, which is in some cases of
greater significance than the enhanced flexural strength.

Centrifugal forming

The flexural strength obtained from experiments is depicted in Fig. II. In respect to the
special consistency of these mixtures, good fiber dispersion and flow properties are
necessary. The results of mix I a and lb show that the increase ofthe w/b ratio influences
only very slightly the flow properties of the mixture and does not improve the fiber
distribution and compaction ability during the spinning. The substitution of the micro-
fine cement through smaller silica fume particles has a positive influence on the fiber
dispersion and mixture flow properties. In addition the use of PVA fibers improves also
the flow of the mixture due to their hydrophilic surface and relative large diameter of
0.1mm, which results in a slight increase of flexural strength (mixes la and 2a).
Additionally, an increase of the spinning speed (mixes 2a and 2b) results in slight
improvements of the compaction and flexural strength. A very strong influence on the
flexural strength llas the addition of a higher amount of latex ( 15 ... 20wt% to binder) (mix
3a, 3b ). Latex is known to have a positive influence on dispersion of carbon fibers, to
lubricate the mix and improve its workability. Better fiber dispersion during the spinning
process was therefore expected. Furthermore, the bond between fiber and matrix is
improved [9). The use on the latex content shows significant differences between these
forming methods (Fig. 12).

CONCLUSIONS

The addition of micro-fine cement to ordinary Portland cement improves the rheological
properties of the fresh mixture and leads to a very dense cement matrix with very high
mechanical properties. A decreasing amount of superplasticizer causes an increase of the
flexural strength, but a decrease in compression. The addition of small volume content
(2vol%) of high performance short carbon fibers enhances significantly the flexural
strength of the composite. of PV A fibers in the mixes results in a decreased flexural
strength (mixes 3c and 3d). This is possibly due to inhomogeneities and segregation
effects.

Finally, a comparison between spun cast and conventionally cast samples in dependence
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

120 Hesselbarth and Kaufmann

Latex dispersion influences positively the fiber dispersion and the fiber matrix bond and
leads to further increases in the flexural properties of the composites. At low doses
(<5wt% to binder), latex influences predominantly the fiber dispersion with an optimum
at 2wt%. A higher amount of latex (15 ... 20wt% to binder) alters the matrix and a limit in
the flexural strength is reached. Fiber type and mechanical properties influence the
efficiency of the fiber reinforcement. The carbon fibers improve predominantly the
flexural strength of the composites, whereas the PYA and PP fibers improve the ductility.
A positive effect of polymer dispersion at high doses (15 ... 20wt% to binder) on the
centrifugation process and on the flexural strength was observed. A further optimization
of the mix proportion and the spinning parameters (spinning speed and time) will be
necessary in order to reach flexural properties similar to those of conventionally cast

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
samples.

ACKNOWLEDGEMENT

The authors would like to thank Dr.G.P. Terrasi and G. Battig for continuous support and
encouragement. This work was supported by SACAC, Lenzburg, Switzerland, and by the
Swiss Commission for Technology under grant no. 4600.1.

REFERENCES

[I] German R., Particle packing charakteristics, Metal powder industries federation,
Princeton, ( 1989).
[2] Bentur A., Mindess S., Fiber reiriforced cementitous composites, Elsevier Applied
Science, England, ( 1990).
[3] Hikasa J., Genba T., Replacement of Asbestos in Reinforced Cement Products
Kuralon, PVA Fibres, Properties, Structure, Proc. Intern. Man-Made Fibres
Congress, Austrian, Chemical Institute, Austria, ( 1986).
[4] Bayasi Z., Development and mechanical characterization of carbon fiber reinforced
cement composites, PhD Thesis, Michigan, (1990).
[5] Larson B.K., Drzal LT., Soroushian P., Carbon-fibre cement adhesion in carbon
fibre reinforced cement composites, Composites, 21 (1990), pp. 205-215.
[6] Soroushian P., Aouadi F., Nagi M., Latex-Modified Carbon Fiber Reinforced
Mortar,ACI Materials Journal, Jan.-Febr. (1991), pp. 11-18.
[7] Burnett E.F.P., Constable T., Cover P., Centrifogated wire fiber reinforced
concrete,ACI SP 44, (1971), pp. 455-475.
[8] Kaufmann J., Winnefeld F., Hesselbarth D., Effect of the addition of ultrafine
cement and short fiber reinforcement on shrinkage, rheological and mechanical

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 121

properties of Portland cement pastes, Cement and Concrete Composites, Special
Issue Early Age Concrete, in press, (2004).
[9] Chen P.W., Fu X., Chung D.D.L., Microstrnctural and Mechanical Effects oj
Latex, Methylcel/ulose, and Silica Fume on Carbon Fiber Reinforced Cement, ACJ
Materials Journal, 94 (1997), pp. 147-155.

Table I : The properties of the fibers

Fiber Young tensile strength diameter (~m] length [mm]

modulus [GPa]
(GPa]

CFNr.l 215 2.5 7 6

CFNr.2 228 3.8 7.2 6.35

CFNr.3 238 3.95 7 6

CFNr.4 30 0.67 18 6
PVA 40 1.6 40 6
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

pp 8.5 0.5 35 X 250~()() 6

Table 2: Composition of the mixtures for centrifugal forming

mix CEMI microfine silica w/b Latex PVA spinning speed

42.5 N cement fume ratio [wt%/ [vol%] [rpm]
[wt%] [wt%] [wt%] binder]
Ia 80 20 - 0.2 2 - 570
lb 80 20 - 0.25 2 - 570
2a 80 - 20 0.29 2.7 2 570
2b 80 - 20 0.29 2.7 2 669
3a 80 20 - 0.2 15 - 669
3b 80 20 - 0.2 20 - 669
3c 80 20 - 0.2 15 2 669
3d 80 20 - 0.2 20 2 669

Table 3: Composition of the mixtures "mix 1'". "mix 2... "mix 3..

no. CEM microtine wlb Super- Lalex pp PYA CF

42.5N «ment ratio plasticizer [wt%/ [vol%] {vol%] no. 2
[wt%] [wt%] [wto/ofbinder] binder] [voll!-0]
I 80 20 0.2 3 0 6.5 0 0
2 so 20 0.2 3 0 0 4.4 0
3 so ltl 0.2 3 2 0 0 2

Copyright American Concrete Institute

122 Hesselbarth and Kaufmann

=
Q..
25
:;
-20
.c
iii
~
..
5 15

-e 10
=~

£ 5

Fig. 1: Flexural and compressive strength of different cementitious matrices with 2

vol% of carbon fibers.

-20 ., ....

llif•.~·:
'! Gj
=4 j· ··················· "'''
-a- 7 days. wfb 0.18 '
--- 28 cla)1s, wlb 0.18
-&--7 da)!s, wlb 0.22 j
L~..~.~~..~~:~.

£ 02 . ;i·····.......,. . ............ ······{

1 1.5 2 2.5 3 3.5

Superplasticlzer (wt"MBinder)

Fig. 2: Flexural strength at w/b ratio of 0.18 and 0.22 (CF 2 vol% ).

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 123

~~r !: .•lIr-.·. . . ...~.···~~·········

14
~~
~~-. ~ . ·. ·. ·. ·. · . · · . ·. ··.
'~>-~--'·"' ~::.~~ . ·.·.·. · .: ~ :--7 clays, wtb 0.18 :
i-ft.- 28 d . . wlb 0.18
~ 81 ~ L·~·V 90days. wlb 0.18;
"! &j l
I :1
~ 0 ·~,--~----~----~
i

1 1.5 2 2,5 3 3.5

Superplastlclzer [wt"le/Blnder)

Fig. 3: Flexural strength at wlb ratio of 0.18 after 7, 28 and 90 days (CF 2 vol% ).

l180
!. 1so I ·~·· . . . /__.. I
l;
~~
.c 140
5
c 120 + . -· ~····-··- ..,
- - 7 days, wlb 0.18
! 100 1I
:
~
80
60 ~
+ I
•-ft.- 28 days, wlb 0.18
.......... 90 clays, 'Nib 0.18
; 40 t
ae
0
20 1
o·~------~~~
. .
(,)
1 1.5 2 2.5 3 3.5
Superplastlcizer [wt"loiBinder]

Fig. 4: Measured compressive strength at wlb ratio of 0.18 (CF 2 vol% ).

CKohlen!ltofffuem
SGL
I:IIKohlenstofffaum
Zoltek
CKohlell$tofffuem
Tenax
OKohlenstofffa!lem
Kureha

7 28 90
Alter [days)

Fig. 5: Flexural strength of various carbon fiber reinforced concrete

compositions (CF 2 vol% ).
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

124 Hesselbarth and Kaufmann

Fig. 6: Fracture surface of cement matrix reinforced with glycerine sized carbon fibers
CFno. I.

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
Fig. 7: Fracture surface of cement matrix reinforced with thermoset sized carbon fibers
CF no. 2 .

...... 34
ca
D. 32 .
!. 30
~
c
28 •• w.•.""'""""""'"'W··"'

-+-7days •
! 26
';j 24 .. ~~~~I)~!
"!22
:s
.; 20
u. 18 .. ·
0 2 4 6 8 10121416182022
Latex amount [wt"lo/Binder)

Fig. 8: Flexural strength dependency on latex amount (CF 2 vol %).

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 125

Mlx1 Mix2 Mlx3

Mixture

Fig. 9: Flexural strength of the composites with different fibers (CF, PVA and PP).

4.5
4.0 / Carbon Fibers

3.5
3.0
2.5
7..
..:.:: 2.0
't:l
=
=
1.5
1.0
"""
0.5
0.0
-0.5
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Displacement JmmJ
Fig. I 0: Load displacement curves (maximum value, 28 days) of the composites with
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

different fibers (CF, PVA and PP) .

..... 30 ,......................................................,...........................................................................................,
10 570rpm 669rpm
!_25 ---
= 20

-e
1:11
15
I ll
"E
...
10
~
Gl
5 l

u:: 0 - '--
_l_ '--· c-... ......_

1a 1b 2a 2b 3a 3b 3c 3d
Mix

Fig. 11: Flexural strength of centrifugated samples (mean values).

Copyright American Concrete Institute

126 Hesselbarth and Kaufmann

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
35~----------------------------~

conventional casting --o

30 0 0
'ei'
~
I~-;---·,.
25
.1:
'Sil 20
=4.1
b... 15
-; ..·• 1a
.
s"":s
10
s 5
specimen fabrication
impossible
ou_~~~~~~-L~~~~~J-~~

0 2 4 6 8 10 12 14 16 18 20 22
Latex amount (wt%/binder)

Fig. 12: Flexural strength casted and centrifuged samples in dependency on the
latex amount.

Copyright American Concrete Institute

SP-224-10

Temperature Controlled Microwave

Accelerated Curing of Precast Ferrocement
Secondary Roofing Slabs

by K. C. G. Ong. C. P. Teo. C. H. Shum. L. H. J. Wong. S. T. Tan.

and C. T. Tam

S):nopsis_;_ The use of microwave technology to speed up the production of precast

ferrocement secondary roofing slabs is explored in this paper. In particular, the use of
discrete on-off microwave curing regimes and the effects of such regimes on the strength
and durability of the ferrocement slabs are investigated. By a regime of on-off microwave
application to maintain the temperature of the slab within a specified range during
microwave curing, overheating of the slabs can be avoided. High early age strengths were
attained in slabs cured using such regimes, with no strength loss at 28 days. In addition,
the durability of such slabs need not be compromised. The use of an appropriate reduced
power level during the later stage of the curing process was found to result in a marginal
improvement in the near surface quality without any reduction in early age strength.
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Keywords: curing; decretized; ferrocement; microwave; temperature

127
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=University of Texas Revised Sub Account/5620001114, User=erur, ert
No reproduction or networking permitted without license from IHS Not for Resale, 01/26/2015 02:27:05 MST
Daneshlink.com
daneshlink.com

128 Ong et al.

ACI Member K. C. G. Ong is Associate Professor and Head, Composites and Protective
Engineering Group, Department of Civil Engineering, National University of Singapore.
A Past President of the American Concrete Institute - Singapore Chapter, he is presently
the Immediate Past President of the Singapore Concrete Institute.

C. P. Teo was a Research Scholar at the National University of Singapore, where she is
involved in exploring the use of microwave technology to speed up the production of
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

precast ferrocement slabs.

C. H. Shum is the Head of Technology Development Section of the Building Technology

Department, Housing and Development Board (HDB). He is a Professional Engineer and
an Accredited Checker.
L. H. J. Wong is currently heading the Technology Development Unit 2 of the Building
Technology Department, HDB. He is involved in research studies conducted at HDB
Prefabrication Technology Center.

S. T. Tan is a Senior Executive Structural Engineer at HDB Prefabrication Technology

Centre, where he is currently conducting research and structural engineering testing
works.

C. T. Tam is a Fellow of ACI and an Associate Professorial Fellow at the National

University of Singapore. His research interest covers concrete materials and technology
with special emphasis on tropical concreting practice and self-compacting concrete.

INTRODUCTION

In Singapore, on-site construction is actively being replaced by precast

construction during the last decade. The economic efficiency of any precast operation is
dependent on a number of factors including the primary conditioning time which is the
time required to attain sufficient strength for demoulding or handling, the total cycle time
or the time from mixing to demoulding, which governs the number of shifts in a day, the
rate of mould reuse, the storage space requirements and energy consumption [1]. The
speed at which the desired properties can be achieved is one of the core requirements for
a successful precast production [1]. It has been estimated that the primary conditioning
aspect of a typical production process makes up as much as 70% of the total cycle time
[I]. A higher rate of strength development could certainly shorten the time needed for the
primary conditioning of these elements and bring about benefits such as an increased rate
of mould reuse and a reduction in idle time. A higher rate of strength development also
facilitates the implementation of just-in-time production whereby the precast elements are
"pulled" through the production system where and when they are needed and thus,
reducing or even eliminating the need for storage space.

The use of conventional heating techniques such as steam curing to accelerate

the strength gain of precast elements has its drawbacks and limitations. Steam curing,
which takes place at atmospheric pressure and at a maximum curing temperature between

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 129

66°C to 82°C, has been found to facilitate the acceleration of cement hydration and
shorten the curing time [2]. Unfortunately, steam curing may result in lower long-term
strength even though high early-age strength can be attained [3]. In addition, steam
curing relies on the thermal conductivity of the specimen. For a material such as concrete
which has very low thermal conductivity, the process will be very slow [1]. Indeed, it has
been reported that within a production run of 13 hours, I 0 hours was spent on the curing
of hollow-core concrete slab elements using traditional heating techniques [4]. Such
heating techniques also require extensive infrastructural set-up and may not be cost-
effective arising from waste in the form of energy losses and low heating efficiency [1].

Microwave technology is believed to have great potential to revolutionize the

curing of precast concrete [5]. Microwave processing of materials makes use of the
internal energy dissipation due to the resistance of molecular dipoles to rotational motion.
This occurs when the microwaves penetrate a material and interact with it,
volumetrically, heating up the material [6]. One feature that distinguishes microwave
curing from other conventional techniques is its ability to heat up a sample very rapidly
due to the deep penetration ofthe microwaves [I]. This translates into a shortening of the
time required for primary conditioning of specimens using microwaves, compared to that
needed using conventional heating techniques. The deep penetration that microwaves can
achieve also allows the specimen to be heated up more uniformly and volumetrically [7].
In addition, this technology has the potential to minimize waste and reduce energy
consumption and hence, results in clean manufacturing [8]. Unlike the long-term strength
of steam-cured concrete, the long-term strength of concrete cured using microwaves can
be higher than that attained for conventionally cured concrete [9].

A collaboration between the Prefabrication Technology Centre (PTC) of the

Housing and Development Board (HOB) of Singapore and the Centre for Construction
Materials and Technology (CCMT) of the National University of Singapore (NUS) has
been in place to look into the production of precast ferrocement secondary roofing slabs
using microwave technology. Precast ferrocement secondary roofing slabs are non-
structural elements provided on the rooftops of public housing flats in Singapore for the
purpose of insulation against solar radiation. In the conventional precast yard, steel
moulds are used to cast the slabs. These slabs are demoulded I day after casting and
moist cured for 3 days. After which, they are air-dried under outdoor conditions until the
time of transportation to the site after 28 days. The use of microwave technology
certainly has the potential to shorten the time spent on the primary conditioning of such
slabs and increase the rate of mould reuse.
One of the critical parameters to control is the temperature of the specimen
during the microwave curing process [7, 10, II]. It has been shown that microwave
curing at extreme temperatures may influence both the macro and micro properties
considerably. High internal temperatures reached (92°C) during microwave curing has
been found to result in the formation of cracks [12]. In addition, slight bulging of the
trowelled surface of mortar cubes has also been observed when the surface temperature
of the cubes reached approximately 98°C at the end of the curing process (13].
Furthermore, curing at temperatures above 70°C at early ages may affect the formation of
ettringite as ettringite may not form at temperatures above 70°C. This may lead to the
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

130 Ong et al.

occurrence of delayed ettringite formation and the possible destructive cracking in
specimens subjected to temperatures above 70°C (14]. As such, there is a need to control
the temperature of the specimen. Leung and Pheeraphan [10] employed feedback
temperature control, which resulted in continuous varying power levels, to limit the
maximum temperature of the specimen during microwave curing. An optimal power
history, identified through feedback power control, was also approximated with a number
of discrete power levels [I 0]. Similar results as those obtained using the continuously
varying power levels were attained (1 0]. By avoiding overheating of the specimens,
Jesser total energy would be required and a more energy efficient process can be achieved
[I 0]. Isothermal heating, which was achieved using feedback temperature control, was
investigated by Sohn and Johnson [7] who reported optimal curing conditions for target
temperatures of 40°C and 60°C. 80°C was, however, found to be unsuitable for
microwave curing [7]. In the studies by Leung and Pheeraphan (10] and Sohn and
Johnson [7], continuous microwave application was employed.

In a study by Teo et al. (12] on the effects of microwave curing on precast

ferrocement secondary roofing slabs, the processing parameters, mainly the power level
and microwave curing duration, were varied by trial and error in a bid to obtain
microwave curing regimes with the maximum temperature reached not exceeding 70°C.
In this paper, the use of discrete on-off microwave curing regimes is explored in order to
limit the temperature reached in the prototype precast ferrocement slab secondary roofing
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

slabs during microwave curing. The influence of such curing regimes on the strength and
durability of the prototype precast ferrocement secondary roofing slabs is also examined.

RESEARCH SIGNIFICANCE

The research reported in this paper deals with the use of discrete on-off
microwave curing regimes (switching on and off when the temperature monitored
reached a specified upper bound or lower bound value respectively) to limit the
temperature reached in the prototype precast ferrocement secondary roofing slabs during
microwave curing. By a series of on-off microwave application once the temperature of
the ferrocement slab reached a specified range, overheating can be avoided. Results
showed that ferrocement slabs of high early age strength and good durability can be
produced with the use of such microwave curing regimes.

EXPERIMENTAL PROGRAM

Objectives

1. To explore the use of discretized on-off microwave curing regimes (constant

power level) in achieving high early age strength and good durability of
prototype ferrocement slabs.
2. To look into the effects of the use of a combination of power levels on the
strength and durability of prototype ferrocement slabs.

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 131

Fabrication of prototype ferrocement slabs

The prototype precast ferrocement secondary roofing slabs studied have

dimensions of 900 mm in length, 600 mm in width and 20 mm in thickness. Each slab
contained reinforcement made up by two pieces of galvanized welded wire mesh (each of
25 mm square grids and 1.5 mm diameter wires) that were lightly spot-welded together
and placed centrally with respect to the thickness of the slab. The slab specimens had
identical slab dimensions and utilized the same type and arrangement of reinforcement as
that currently in use. All ferrocement slab specimens were cast in fibre-glass moulds as
they have good thermal stability and are transparent to microwaves. Figs. l(a) and l(b)
show the plan and sectional views of the prototype ferrocement slab.

In this study, ordinary Portland cement (OPC) and natural sand were used. The
mortar had a water-cement ratio by weight of 0.45 and a sand-cement ratio by weight of
2.0. The mortar was mixed in a container for I 0 minutes using an electric handheld
mixer. After placing the mortar into the moulds, the slab specimens were vibrated on a
vibrating table to achieve adequate compaction.

:rvt icrowave Curing of Specimens

The same prototype mechanized industrial microwave curing system (Fig. 2),
developed by the Prefabrication Technology Centre in Singapore, was used throughout
this study. This microwave curing system is capable of a maximum power level of 6 k W
and generates 2.45 GHz fixed frequency microwaves. Output power can be controlled
from 10% to I 00% of the maximum power level for various time intervals as desired. For
this curing system, microwave penetration into the specimen occurs predominantly from
the top, trowelled surface.
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

The 0.75 m3 curing chamber is capable of accommodating up to three (900 X

600 x 20) mm ferrocement slab specimens. In this study, only one specimen, however,
was cured in the curing chamber at any one time. A built-in conveyor belt system
provides the necessary access for the specimens into the curing chamber for microwave
curing and out of the chamber after microwave curing. A microwave leakage detector
was also used to ensure that the curing system conforms to international standards and
that it does not leak microwaves greater than 5 m W/cm 2 at a distance of 50 mm from any
surface.

In this study, discrete on-off microwave curing regimes were used whereby the
microwave curing system was switched off manually when the temperature at a location
(amongst the locations chosen for temperature monitoring) reached the upper value of a
specified temperature range and switched on again manually when the temperature at all
the locations monitored were lower than or equal to the lower value of the specified
temperature range. This continued for a specified duration. In this way, the temperature
of the slab specimen, at all the locations monitored, would not exceed the upper value of

Copyright American Concrete Institute

132 Ong et al.

the specified temperature range during microwave curing. Such microwave curing
regimes are used here to simulate feedback temperature control processes. The variables
used are the microwave power level, the specified temperature range and the specified
on-off curing duration. All the slab specimens were covered with a microwaveable film
during the microwave curing process.

The designation used in this study for a slab specimen subjected to a particular
microwave curing regime is as such: (20)/lh/(60-55)-3 refers to the slab specimen which
was subjected to a microwave power level of 2 kW and having a delay time of 30 min.
When the temperature monitored on the slab reached 60°C, the curing system was
switched off and the slab specimen was allowed to cool down inside the curing chamber.
When the temperatures at all the locations monitored were lower than or equal to 55°C,
the curing system was switched on again. This continued for a total duration of 1 h.

Though the majority of the specimens were subjected to a constant power level
throughout the curing process, the use of a combination of power levels during
microwave curing was also investigated. For slabs subjected to such regimes, a specific
power level was used in the first half of the on-off curing duration. This was followed
with the use of a different power level in the second half of the on-off curing duration.
For example, slab specimen (30-06)/1 h/(65-60)-4 was first subjected to a power level of 3
kW after a delay time of 40 min. The specified temperature range was 60 oc to 65 °C.
After 30 min, the power level was lowered to 600 W and this power level was used for
the second half of the curing duration.

Temperature Monitoring

Unshielded type 'T' thermocouples were directly inserted into the specimens to
chart the temperature variation with time of the slab specimens. These thermocouples
were placed at 5 different locations (denoted as 1, 2, 3, 4 and 5) and at 2 different depths
(either near the top, exposed surface (T) or near the bottom, mould face (B)). The
locations of the embedded thermocouples are shown in Fig. 1(a). In a companion paper, it
was found that curing using the present microwave curing system produces spatial
temperature non-uniformity, with the temperature of the slab near the edges registering a
higher rate of heating and a larger temperature rise as compared to areas near the centre
of the slab [12]. As such, locations near the edges of the slab were chosen for temperature
monitoring in an attempt to limit the temperature within the entire slab to a specified
temperature range. Temperature measurements were recorded with the use of a data
logger at intervals of one minute. The type 'T' thermocouples used can measure
temperatures ranging from 0°C to 350°C with an accuracy of± 1°C.

The temperatures monitored by the thermocouples are denoted as such: the

number refers to the location of temperature monitoring while the subsequent letter refers
to the depth at which temperature was monitored. For instance, '1 T' refers to the
temperature of the mortar near the top, trowelled surface monitored at location 1.

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 133

Initial Surface Absorption Test (ISAT)

The initial surface absorption test is one of the quality control tests that have
been used for conventional precast ferrocement secondary roofing slabs. This test is
described in BSI881: Part 208: 1996. It is a low-pressure assessment of the water
absorption of the mortar and gives an indication of the near surface quality and hence, the
durability, of the ferrocement slab specimens. The rate of flow of water per unit area into
the specimen through the surface at 10 min, 30 min and 60 min from the start of the test
was obtained. The entire test was carried out under a constant applied head of (200 ± 20)
mm and performed on the trowelled face of the ferrocement slab specimens.

For !SAT, the prototype ferrocement slabs were demoulded a day after casting.
The slabs were then moist cured by spraying with water and covering with wet hessian
for three days. Following which, the slabs were air-dried under indoor ambient laboratory
conditions till the lSA Twas performed at 28 days.

Patch Load Test

To assess the early-age and later-age strength of ferrocement slab specimens

subjected to the various microwave curing regimes, the patch load test was employed.
This test is one of the quality control tests that are being used for conventional precast
ferrocement secondary roofing slabs. The ferrocement slab specimens were simply
supported, unrestrained at all four corners, and statically loaded to failure under a central
patch load of size (I 00 x I 00) mm. In service, the conventional precast ferrocement
secondary roofing slabs are placed with the mould face facing upwards. Hence, the
ferrocement slab specimens were placed and tested in a similar orientation to simulate
actual loading conditions.

Slabs were tested at 4.5h and at 28 days. Slabs subjected to patch loading at 4.5h
were demoulded just prior to testing. Slabs which were subjected to patch loading at 28
days were demoulded I day after casting and moist cured for a further 3 days under wet
hessian. Thereafter, they were air-dried under indoor ambient laboratory conditions till
the time of patch load test.

All slabs subjected to the patch load test were cut along the failure section. The
actual position of the reinforcement as well as the thickness of the slabs at the failure
section were measured with the use of a venier caliper. The theoretical load carrying
capacity of each ferrocement slab specimen under central patch loading was then
computed from the slab thickness, the effective depths of the reinforcement, properties of
the reinforcement as well as the appropriate compressive strength of the mortar.

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

134 Ong et al.

RESULTS AND DISCUSSION

Temperature distribution, initial heating rates and energy consumed

Fig. 3 shows the typical temperature variation with time of ferrocement slab
specimens subjected to discrete on-off microwave curing regimes at constant power
levels. From the temperature profiles, information on the total energy consumed as a
result of microwave curing, the initial heating rate, and the heating and cooling pattern
can be obtained for individual slab specimen. The typical heating-cooling patterns for
slab specimens subjected to such microwave curing regimes are shown in Table I. In
general, the number of on-off curing cycles increased with increasing microwave power
level. This is due to the rapid heating that is associated with a higher microwave power
leveL

Fig. 4 shows the initial heating rates for the slab specimens subjected to the
various power levels and temperature ranges. These values were calculated using the
average temperature rise registered at the hottest location during the first cycle of
microwave application. It can be seen that the initial heating rates increase with an
increase in power level. In addition, it was found that the variance in the initial heating
rate increased with increasing power level. This implies that the lower power level results
in a more consistent temperature profile. Fig. 5 also shows that for the same power level,
the use of a lower temperature range produced a higher initial heating rate. This can be
attributed to the use of the average temperature rise of the hottest location (amongst the
locations monitored) for the computation of the initial heating rate and the finding that
the rate of temperature rise decreases with increasing microwave curing duration.
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

The energy consumed by the ferrocement slabs as a result of exposure to

microwaves is shown in Fig. 5. Generally, more energy is required when a longer curing
duration or a higher temperature range is used, or a combination of both. Regimes with a
2h on-off curing duration and a temperature range of 60 oc to 55 °C required the most
energy due to the long curing duration and high maximum temperature. For slabs
subjected to the same on-off curing duration and specified temperature range, it appears
that the power level does not have a significant influence on the total energy consumed as
slabs subjected to a higher power level were exposed to the microwaves for a shorter
period of time and vice versa such that the influence of power level on energy consumed
is minimal. In addition, it seems that the temperature range used has a greater effect on
the amount of energy required than the on-off curing duration. An increase in temperature
range used from 45 oc to 50 oc to 55 °C to 60 °C meant an increase in average energy
needed of approximately 54% whereas an increase in on-off curing duration from 1 h to 2
h only resulted in an increase in average energy required of 18%.

4.5h ultimate load carrying capacity

Prototype ferrocement slabs, subjected to the various microwave curing regimes,

were tested at 4.5 h under the patch load test. The results are presented in Fig. 6. All slabs
failed in a flexural mode. The strength of the mortar of the ferrocement slabs at 4.5 h was

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 135

not sufficient to result in the yielding of the reinforcement and consequently, the slabs
failed due to crushing of the mortar at the top of the slab which occurred soon after
cracks appeared at the bottom of the slab. To compare the results for the slabs subjected
to the various regimes, it is most informative to consider the normalized strengths,
obtained using the normal cured slab as the base, to eliminate interference from the slight
variation in the effective depths of the reinforcement of the slabs. It can be seen that the
microwave cured slabs failed at ultimate load values that were much higher than that of
the normal cured slab. This indicates that microwave curing has the potential to enhance
the early age strength of ferrocement slabs. Higher 4.5h strengths were obtained with the
use of a longer on-off curing duration or a higher temperature range. The use of a 2h on-
off curing duration coupled with the higher temperature range of 55 oc to 60 °C resulted
in the highest strength gain, with the 4.5 h strength of slabs subjected to this combination
of curing duration and temperature range being approximately 10 times that of a normally
cured slab. This seems to indicate that 2 h/(60 oc to 55 °C) is the most optimal
combination of temperature range and curing duration used, in terms of early age strength
gain.

28-day uliimate load carrying capacity

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

The theoretical and experimental 28-day ultimate load carrying capacities of

prototype ferrocement slabs are presented in Table 2. All slabs failed in flexure with high
ductility. The theoretical values were based on yield line analysis of the slabs supported
on four unrestrained corners and took into account the actual yield strength and position
of the reinforcement, the thickness of the slab and the appropriate 28-day cube
compressive strength of the mortar. The experimental ultimate loads of slabs subjected to
the various discretized on-off curing regimes at constant power level and having a delay
time of 30 min ranged from 2.9 kN to 4.3 kN. It can be seen from Table 2 that the
experimental ultimate load values were higher than their respective theoretical values. A
typical factored ultimate load requirement specified by HDB for such slabs is 2.35 kN as
such slabs are used as non-structural elements when they are placed on the rooftops of
public housing flats [5]. All the slabs tested satisfied this strength requirement.

Initial surface absorption of slabs subjected to discretized on-off curing regimes

Fig. 7 shows the ISA T results for tests conducted on the trowelled face of the
prototype ferrocement slabs. It can be seen that the majority of the slabs passed the
requirement specified by HDB as well as that suggested by Levitt [ 15]. Slab specimens
(20)11 h/(60-55)-3 and (30)/1 h/(60-55)-3 had very high ISA T values. Although all slab
specimens were covered with a microwaveable film before, during and after microwave
curing, the film for these two specimens could have come off the surface during
microwave curing, leading to a significant amount of moisture loss from the surface of
the slab at the high temperature of 55 oc to 60 °C and a poorer near surface quality. It is
noticed that slab specimens (20)/1 h/(50-45)-3 and (30)/1 h/(50-45)-3 had ISA T values
which were slightly higher than the !SAT values specified by HDB. Due to the relatively
higher power level and lower temperature range used of 45 oc to 50 °C, these two
specimens ((20)11 h/(50-45)-3 and (30)/1 h/(50-45)-3) were exposed to microwaves for

Copyright American Concrete Institute

136 Ong et al.

only a short period of time (12 min and 8 min respectively). Hence, the amount of
moisture loss might be smaller for these slab specimens, leading to a higher resultant
water-cement ratio and contributing to a slightly poorer near surface quality of the
mortar.

Effect of the use of a combination of power levels on the strength and near surface
quality of ferrocement slabs

A typical temperature vanatJon with time for slab specimens subjected to

variable power levels is shown in Fig. 8. Typical heating-cooling patterns for such
specimens are shown in Table 1. In this study, a higher power level
(3kW) in the first half of the on-off curing duration followed by a lower power level
(either IkW or 600W) in the second half of the on-off curing duration was used. As
reported by Leung and Pheeraphan [10], a more rapid heating during the early stage of
the curing process appears to be closer to the optimal process. Using a lower power in the
second half of the curing process also means that the slab specimen is subjected to a
lower heating rate during the later stage of the process. Slab specimen (30)/Ih/(65-60)-4,
which was subjected to a constant power level, was also cast to serve as a basis for
comparison.

The 4.5h ultimate patch load capacities of slabs subjected to variable power
levels are shown in Fig. 9. There is no significant difference in terms of early age strength
for (30)11 h/(65-60)-4 and (30-1 0)/1 h/(65-60)-4. However, it appears that the use of a
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

much lower microwave power level during the second half of the process (as is the case
for (30-06)/Ih/(65-60)-4) is insufficient to induce as rapid an early strength gain as the
use of a higher power level (I kW or 3kW).

The 28-day theoretical and experimental ultimate patch load capacities of slabs
subjected to variable power levels are shown in Table 2. The theoretical values were
calculated in a similar manner as that for the 28-day theoretical load values of slab
specimens subjected to constant power levels. The experimental ultimate load values for
slabs subjected to variable power levels were also found to be higher than their
corresponding theoretical load values and failed at ultimate load values which were
higher than the required factored load value.

The ISA T results for the set of specimens subjected to variable power levels are
presented in Fig. 10. A comparison of the ISAT values for (30)/Ih/(65-60)-4 and (30-
IO)/Ih/(65-60)-4 seems to indicate that an appropriate reduction of power level in the
second half of microwave curing (from 3kW to I kW) may lead to a slight improvement
in near surface quality of mortar. The use of a very low power level of 600 W in the
second half of microwave curing could lead to a lower amount of free water loss, leading
to a poorer near surface quality of the mortar.

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 137

CONCLUSIONS

The use of discretized on-off microwave curing regimes in achieving high early
age strength and good durability of prototype precast ferrocement slabs is explored in this
study. Results showed that such curing regimes can be used for microwave curing of
ferrocement slabs .. By limiting the temperature reached in the slabs, the occurrence of
extreme temperatures and overheating can be prevented. The microwave cured slabs
showed much higher 4.5h patch load strength when compared to normal cured slab,
without any strength loss at 28 days. In addition, durability of the ferrocement slabs
subjected to discretized on-off microwave curing regimes need not be compromised. The
use of an appropriate reduced power level during the later stage of the microwave curing
process can lead to a marginal improvement in durability of the slab with no significant
difference in the early age strength.

ACKNOWLEDGEMENTS

The authors gratefully acknowledge the assistance of Mr. K. S. A. Chew in the

carrying out of the study reported in this paper.

REFERENCES

( 1) Mak, S. L. Microwave Accelerated processing for precast concrete production.

In Proc. Fourth CANMET/ACI International Conference on Durability of
Concrete, August 1997, Sydney, Australia, pp. 709-720, Supplementary papers.

(2} Alunno-Rossetti, V., G. Chiocchio and M. Collepardi, Cement and Concrete

Research, Vol. 6, pp. 279-298. 1974.

(3} Alexanderson, J.. Behaviour of concrete under temperature extremes, SP 39-7,

ACI Special Publication, SP39, Detroit, Michigan. 1973

(4) Bella, S. B., Lai, S. and M. Pinna. Microwaves for the hyperaccelerated curing
of concretes, Betonwerk + Fertigteil-Technik., Issue 12, I 994.

(5) Lau, J. M., Tan, K. B., Oh, L. S., Tan, C. K., Ong, K. C. G. and S. Sabesan.
Microwave accelerated production of ferrocement slabs- An industrial
perspective. In Proc. Fifth CANMET/ACI International conference on Recent
advances in concrete technology, July-August 200 I, Singapore, pp.5 I 7-535.

(6) Thostenson, E.T. and T.-W. Chou. Microwave processing: fundamentals and
applications, Composites: Part A Applied science and manufacturing, Vol. 30,
No.9, pp. 1055-1071. 1999.

(7) Sohn, D. and D. L. Johnson. Microwave curing effects on the 28-day strength of
cementitious materials, Cement and Concrete research, Vol. 29, pp. 241-247.
1999.
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

138 Ong et al.

(8) Mak, S.L. A quantum leap. 1998.
http://www .dbce.csiro.au/innovation/1998-1 0/concrete.htm

(9) Wu, X., Dong, J. and M. Tang. Microwave curing technique in concrete
manufacture, Cement and concrete research, Vol. 17, pp. 205-210. 1987.

(10) Leung, C. K. Y. and T. Pheeraphan. Determination of optimal process for

microwave curing of concrete, Cement and Concrete research, Vol. 27, No. 3,
pp. 463-472. 1997.

(11) Mak, S. L., Shapiro, G. and T. Son. Accelerated heating of concrete with
microwave curing. In Proc. Fourth CANMET/ACIIJCI International conference
on recent advances in concrete technology, June 1998, Tokushirna, Japan, pp.
531-542.

(12) Teo, C.P., K.C.G. Ong, C.H. Shum and S.T. Tan. Accelerated heating of precast
ferrocement secondary roofing slabs using microwave energy. In Proc. 2ih
Conference on Our World in Concrete and Structures, 29-30 August 2002,
Singapore, Vol. XXI, pp.589-596.

(13) Sabesan, S.. Microwave curing of precast slabs. M.Eng. Thesis, National
University of Singapore. 2001.

(14) Taylor, H.F.W, C. Famy and K.L. Scrivener. Delayed ettringite formation,
Cement and Concrete research, Vol. 31, pp. 683-693.2001.

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
(15) Levitt, M. The ISAT- A non-destructive test for the durability of concrete,
British Journal ofNDT, pp. 106-112. July 1970.

List of notations

P u, theory Theoretical load carrying capacity

Pu, expt Experimental load carrying capacity

Conversions for unit of measurements:

1 in = 0.0254 m = 25.4 mm
I kip-force= 4448 N
temperature (F)= 1.8[temperature (°C )] + 32

Copyright American Concrete Institute

Provided by IHS under license with ACI Licensee=University of Texas Revised Sub Account/5620001114, User=erur, ert
No reproduction or networking permitted without license from IHS Not for Resale, 01/26/2015 02:27:05 MST
Daneshlink.com
daneshlink.com
No reproduction or networking permitted without license from IHS
Provided by IHS under license with ACI
Copyright American Concrete Institute

Table l: Typical heating-cooling patterns l(lr slab specimens subjected In discrctizc(l on-off microwave curing regimes

Slab specimen Energy Total'on' Initial rate

designation consumed time of heating Heating-cooling pattern Sum
. . . . . . . . .. ....... . . . .. .. (~~~t.·~.i.~L.....(~t.•ir•t....ef:!~~.IJ.!l ......... . ........... ......... .............J'!l.~~.L.................. ... .
············~····· ..('l.~.it.•l.
(30)/2h/(60-55)-3 45.0 15 263 7 3 I 4 7 F 9 1 10 T 12!16 132 I 12 120
(20)/2h/(60-55)-3 50.0 25 187 10 2 2 3 .4 2 9 2 10 2 13 2 21 2 30 ~. 3 120
-f
...J.l())~hJ~0-5_?)·3____ _j?.:Q_ _
(30)/2h/(50-45)-J 30.0
__±?___,.....1§.:..2_
10 J.W
33 ?
4 7
194
II >I
34
l3
------
J 18 1 23 1 30
120
120
=-
-·=
(20)/2hi(50-45)-3 28.0 14 207 6. 7 2 18 >2 24, 2 33 2 24 120
=
(10)/2h/(~0-45J-.l
(30)!Ih!(60-55)-3

...
(20)/lh/(60-55)·3
37.0
36.0
40.0
JlO)ltiJI<60_~.?~l:L__ ... 44.,.o
37
12
20
77
240
151
17 7 4
>7 4 2
.l2 3 2
34 . ........... _sL_~LL1 9 ····~ ....... ___ _
9 4 14 4
10 1. 9 I
(> 2.
8
18
13
II
4
1
2
24 4 11
12
12
120
60
60
60
-
5!.
=
...n
=
CD
Not for Resale, 01/26/2015 02:27:05 MST
Licensee=University of Texas Revised Sub Account/5620001114, User=erur, ert

CL.
(30)/lh/(50-45)-3 24.0 8 464 J 2 J . • 5 Ll 9 I 12 l, 15 l 9 60
C"')
(20)/lh/(50-45)-J 24.0 12 198 6 7 2 14 2 22 2 5 60 CD
(10)/lh/(50-45)-3 23.0 23 75 15 II 4. 15 4 II 60 3
CD
(30)/lh/(65-601-4
(JO-IO)!lh/(65-60)-4
54.0
45.0
18
19
194
234
Tl
9
7 2 8 2 8
3 2 5 ··I 8
9
101615 •
10 l . 60
60
=
"':"
CCI
Change in power level from I»
3k\Vto I kW
en
CD
(30-06)/1 h/( 65-(>0)-4 51.6 34 2!5 JQ 3 5 6 7121 5
Change in power level from
60
.,...
CL.

3kWto600W =
--
=
CL.
n
Daneshlink.com

microwave on
en

w
ca
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
daneshlink.com

140 Ong et al.

Table 2: 28-day theoretical and experimental load carrying capacities (denoted as
Pu,theory and Pu,exp1 respectivcly) offcrrocement slab specimens subjected to
various discretized on-off microwave curing regimes

Pu,expt
Slab specimen Pu,theory Pu,expt
designation (kN) (kN) Pu,theory

SNc* 1.25 3.30 2.65

(30)/1h/(60-55)-3 1.68 3.40 2.02
(20)/lh/(60-55)-3 1.76 4.30 2.44
( 10)/1h/(60-55)-3 2.00 4.00 2.00
(30)/lh/(50-45)-3 1.68 3.90 2.32
(20)/lh/(50-45)-3 1.44 3.50 2.42
(I0)/1h/(50-45)-3 1.61 3.60 2.24
(30)/2h/(60-55)-3 1.56 2.90 1.86
(20)/2h/(60-55)-3 1.83 3.50 1.91
(I 0)/2h/( 60-55)-3 1.76 3.40 1.93
(30)/2h/(50-45)-3 1.30 3.00 2.31
(20)/2h/(50-45)-3 1.65 3.70 2.24
(I0)/2h/(50-45)-3 1.61 3.50 2.17
(30)/1 hi(65-60)-4 1.38 2.96 2.15
(30-06)11 h/( 65-60)-4 1.35 3.12 2.31
(30-1 0}11 h/(65-60)-4 1.23 2.99 2.43

• Designation for normal cured slab

Direction of entry
into curing chamber
C>
lA. --``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

X
600 2
mm

............... x1 ..................................................................................;x5!..l
1 1
( [A 900 mm )

Fig. l(a): Plan view of ferrocement secondary roofing slab showing locations of
embedded thermocouples
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=University of Texas Revised Sub Account/5620001114, User=erur, ert
No reproduction or networking permitted without license from IHS Not for Resale, 01/26/2015 02:27:05 MST
Daneshlink.com
daneshlink.com

Thin Reinforced Cement-Based Products 141

-J,25 mm
~--n=~t=

600
mm --3> ¢0mm

10m 1~ -J,25 mm
-----
--3>/ I~ 1'
30mm
Fig. I (b): Sectional view of ferrocement secondary roofing slab (section A-A)

Fig. 2: The prototype mechanized industrial microwave curing system used and the
temperature monitoring of a slab specimen using a data logger

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

142 Ong et al.

65 I~E--~00~--+~~~~
601
- 5Sj
~ .
fso
B
i 45

!40
35

10 20 36 40 SO 60 7iJ 86 90 JOO JlO 120 130

Time from start of mkrowave cur!og (min)

Fig. 3:Typical temperature variation with time for slab specimens subjected to constant
power level [plotted for slab specimen (l0)/2h/(60-55)-3]

:2
:~-~
J"'
__ Mmgtl•h~IO.a"""
tf'inptniiUI11~
..,, .,..,,, P'M'~''""" I I

I
t
.e ,.
~

c
300

zso ,,
i !i
,c
,,
I·'

.. ~
~
,"'
>I
;g I
"
il i

"'
li
!;
\(
.,
II I
3\i\\1 JJt\\'1 lt.WI lkW! 1 k\\'j I k\\l
60-~tt: !IO..t5"'C 60-~!i"C 5&4~"C tie-~ '""IS"<'

Fig. 4: Initial heating rates for ferrocement slabs subjected to various microwave power
levels and temperature ranges

Fig. 5:Energy consumed by the ferrocement slabs as a result of microwave curing

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 143

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Fig. 6:Nonnalized 4.5h patch load strength of ferrocement slabs subjected to various
microwave curing regimes

. ,,.,., .•.. --- -.-

--- _ , . , . . .
-+-i>O>/l ..\ ...$S}J
..... ,.,.,,1~)-J
-----
....,_(!Q)IIM(,G.S.>l
...,_.(IO)"IM6i>U).J
-+-(,.YIM;$V.<S~l

......... ,,O},..
'(~S)<J ---(30>~1$~
i'u.)s , ........ (/0)'l!<\6i>ml ..._ nO>'ll>~...ss)o)
~~~ .
---~~ .. , -t:o~~)-.11
!_O~·
-e-(n>)aj~l _....,.(~ 151
lHW·
i
~-~.21>

~0.15.
l.I.IO

Fig. ?:Initial surface absorption of ferrocement slabs subjected to various microwave

curing regimes

-l·T -)·T
····.f.T -S.T

w ~ ~
D ~ •
'
'
-
00 ~
Time from •tart ofmlt"""""'t tllrfAa (mm)

Fig. 8:Typical temperature variation with time for slab specimens subjected to variable
power levels [plotted for slab specimen (30-06)/lh/(65-60)-4]

Copyright American Concrete Institute

144 Ong et al.

l.S -,.----.!:0::4::,.5::..hn!!!'onna=l:=ised"'-":soreng'?'::''th'---,--';X'-'en"'nergy'-7'•<:: :ons::=:um:=,=.ed::.:by::L.::slab=--,.
l 60
-- -4.5h ultimate pa!Ch load ~tmtgth of normal cu,.d slab
1.6
so
1 1.4
~ i
:- 1.2
- 40 ~

~-~
0!,
:;I.o
•r. 30 1

t-
.c a
::1
...
~0.8 i
0.6 20 l
z: 0.4
10
0.2
······················--------------------··························-··---- -----------········
0.0
(30)/lh/(65-60)-4 (30-IOYlh/(65-60)..4 (30-06)/lh/(65-60)-4
Slab specimen denotation

Fig. 9: Effect of the use of a combination of power levels on the normalized 4.5h patch
load strength of ferrocement slabs

0.12
- (30)/lh/(65-60)-4

"~::.
0.10 ~ (J0-10Ylhl(65-60)..4

...... (30-06)/lh/(65-60)-4
! 0.08
"
0
-:
- - HOB's specification

~006
.....
.!
0.04
-t
ill
ii
.s=
O.Q2

0.00
10 20 30 40 50 60 70
Time from start of ISAT (miD)

Fig. 10: Effect of the use of a combination of power levels on the near surface quality of
ferrocement slabs

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

SP-224-11

Freeze-Thaw Durability of Commercial

Fiber-Reinforced Cement Board

by K. G. Kuder and S. P. Shah

Synopsis: Fiber-reinforced cement board (FRCB) is increasing in consumer popularity

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

because it is more durable than conventional wood products. However, concerns exist
about the freeze-thaw durability of the material due to its laminated structure and high
porosity. To overcome these weaknesses, some manufacturers have begun to press the
material after it is formed. The objective of this work is to evaluate the effects ofthis new
processing on the durability of the FRCB. Three commercially-available FRCB products-
two that had been pressed and one that had not - were subjected to accelerated freeze-
thaw cycling according to a modified version of ASTM Standard C1185. The flexural
strength, interlaminar bond (ILB) strength and porosity were measured. The results
indicate that pressure might improve the ILB and flexural strength ofthe FRCB after
freeze-thaw testing. However. porosity is not affected by pressure after freeze-thaw.

Keywords: fiber·reinforced cement board; freeze-thaw durability;

Hatschek process; interlaminar bond strength; porosity; pressure

145
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=University of Texas Revised Sub Account/5620001114, User=erur, ert
No reproduction or networking permitted without license from IHS Not for Resale, 01/26/2015 02:27:05 MST
Daneshlink.com
daneshlink.com

146 Kuder and Shah

Katherine G. Kuder is a graduate student at Northwestern University, where she
received her MS in Civil Engineering. She received an ACI Student Fellowship for the
2002-2003 academic year. Her research interests include fiber-reinforcement, durability
of concrete, the extrusion process and the use of cementitious materials for residential
construction.

Surendra P. Shah is a Walter P. Murphy Professor of Civil Engineering at Northwestern

University and the director of the Center for Advanced Cement-Based Materials. His
research interests include fiber-reinforced concrete, nondestructive testing and impact and
impulsive loading. He has published more than 300 technical articles and has edited and
co-authored numerous books.

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
INTRODUCTION

FRCB has been used in the United States building industry since the 1980's for
residential applications such as siding, roofing tile, backerboard and fences (1 ). It was
the fastest growing market in the 1990's and by 2005, fiber-reinforced cement might gain
up to 25 to 30% of the siding market (2). The material is increasing in consumer
popularity because it is more fire-resistant, can better withstand fading, is not susceptible
to insect attack and is more durable than conventional wood products.

Fibers are added to cement because of their crack-arresting ability, which

increases the toughness of naturally brittle cement products. The toughness of the
material is enhanced by its ability to undergo multiple cracking before the cracks
coalesce into large macrocracks (3,4). Processing techniques, such as the Hatschek
process, can be used to manufacture thin sheet products with large percentages of fibers
easily and cost-effectively, thus providing an innovative material to the construction
market.

FRCB are thin sheet, laminated elements that are produced by the Hatschek
process and autoclave-cured. The material consists primarily of silica, cement, cellulose
fibers and water. Figure 1 shows the dry composition of the FRCB, by weight.

Figure 2 presents a schematic of the Hatschek formation machine. A dilute

slurry of cement, silica and cellulose fibers exist in the bins. Sieves rotate through these
bins, gathering thin monolayers of the material and placing them on the conveyor belt
where excess water is removed by a vacuum. The monolayer then continues to the next
bin, where another monolayer is added, and so on. This grouping of monolayers is
designated as a layer of the material. These layers are gathered on the accumulator roll
until the desired board thickness is reached. The FRCB then proceeds down the
manufacturing line and is autoclave-cured. The resulting product is reinforced with a
significant amount of cellulose fibers, 30-35%, by volume, and is composed of at least
three layers, each layer consisting of approximately 7-9 monolayers. The laminated
structure ofthe board is shown in Figure 3.

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 147

Freeze-thaw action can cause severe deterioration to external building products
due to the hydraulic pressures that develop as water freezes. Upon freezing, water
expands to a volume 9% greater than its initial volume, exerting tensile forces on the
material. Once these forces exceed the strength of the material, cracking occurs, causing
irreversible damage. During subsequent thawing, the water moves through the cracks,
expanding them further, and is present to cause more damage when freezing occurs
agam.

Numerous studies have been conducted to evaluate the freeze-thaw durability of

cementitious materials reinforced with cellulose fibers. After exposure to natural aging
conditions, an increase in both flexural strength and stiffness was observed due to the
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

petrification of the fibers and an increase in the fiber-matrix bond (5-10). The dynamic
modulus increased for wood-fiber reinforced cast composites that were exposed to
accelerated aging, due to the densification of the material (11 ). Both moisture absorption
and porosity were found to be indicative of freeze-thaw durability. Materials with high
moisture absorption capacities were less durable (12). Porosity studies indicated that if
the pores were either small enough to limit the ingress of water, or large enough to allow
the flow of water and the expansion of ice during freeze-thaw cycling, then Jess damage
would occur ( 13).

FRCB is susceptible to freeze-thaw damage for three reasons. First, it is a

laminated material, providing paths of inherent weakness for water to ingress (12). Once
present between the layers, water will expand during freezing causing the material to
delaminate. Second, it is highly porous, allowing large amounts of water to infiltrate,
which will expand during freezing, leading to cracking. Finally, the material is
reinforced with a significant amount of cellulose fibers. Since cellulose fibers are
organic, their properties may change with time, thus altering the properties of the material
they reinforce. In addition, cellulose fibers are hydrophilic, so they may absorb
significant amounts of water, which weakens the fiber-matrix bond (14,11). During
freezing and thawing these fibers might shrink and swell, further weakening the fiber-
matrix bond and, hence, the mechanical properties of the FRCB.

In an attempt to overcome these weaknesses, some manufacturers have added a

step in the FRCB production process. After the board is formed, it is pressed, expelling
excess water and possibly decreasing porosity and improving the interlaminar bond.
Previous research has indicated that pressing does improve the performance of fiber-
reinforced cementitious sheets before exposure to freeze-thaw conditions. Both density
and flexural strength were improved when asbestos-reinforced cementitious sheets were
pressed (15). Wood-reinforced cement boards that were pressed after casting showed a
reduction in porosity and an improved interfacial bond ( 16, 17).

While the previous research provides useful inforn1ation about the durability of
cellulose-reinforced cementitious materials, applying the work to FRCB is difficult due to
its unique characteristics. Freeze-thaw tests need to be conducted on actual
commercially-produced Hatschek materials to truly understand the material's durability.
Work has shown that pressure improves the properties of cellulose fiber-reinforced

Copyright American Concrete Institute

148 Kuder and Shah

cementitious materials before freeze-thaw, but the effects after freeze-thaw have yet to be
determined.

The objective of this research is to evaluate the effects of pressure treatments on

the freeze-thaw durability of FRCB by comparing three commerciaJly-available FRCB.
The mechanical and microstructural properties of the boards are evaluated both before
and after freeze-thaw cycling. The role of the laminated structure, high porosity and
cellulose reinforcement in the freeze-thaw resistance of the material is determined.

RESEARCH SIGNIFICANCE

The use of FRCB in residential applications continues to increase. With its

superior durability, it provides consumers with an excellent alternative to conventional
materials. However, if FRCB fails in freeze-thaw intensive climates, delaminating in an
unsightly manner, consumer confidence wiJI be lost, and so will the residential
construction market. Some manufacturers have begun pressing the FRCB in an attempt
to improve the freeze-thaw resistance and, thus, reduce the likelihood of delamination
failures. The results indicate that pressing the FRCB may increase the material's
resistance to delamination after freeze-thaw and, therefore, may be an improvement to the
FRCB production process.

EXPERIMENTAL WORK

Materials
Three commercially-available FRCB materials were investigated. All materials
were unsealed 8 mm (5/16 in) thick, Hatschek produced FRCB with cellulose fibers that
had been autoclave-cured. Materials A and B were pressed, Material C was not pressed.
Neither the magnitude, nor the method of application of the pressure treatments, is known
for Materials A and B. In addition, the exact composition of the FRCB, as well as the
details of the formation process, are unknown for all of the materials.

Freeze-Thaw Cycling
Materials were subjected to as many as 300 accelerated freeze-thaw cycles
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

according to a modified version of ASTM C 1185 "Standard Test Methods for Sampling
and Testing Non-Asbestos Fiber-Cement Flat Sheet, Roofing and Siding Shingles, and
Clapboards (18)." Cycling was conducted in a Humbolt freeze-thaw chamber. Saturated
samples were subjected to temperatures ranging from 20°C to -20°C. Each cycle took
approximately 12 hours. -

Mechanical Properties
Flexural performance, including strength and toughness, and ILB strength were
evaluated for each FRCB, both before and after freeze-thaw conditioning. Three
replications were made for each parameter evaluated.

Flexural performance was determined using closed-loop testing by a three-point

bend test. Specimens were 175 mm x 50 mm x 8 mm. The span was 152.4 mm and the

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 149

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

displacement control rate was 0.00381 mm/sec. Load and deflection data was recorded
and converted to obtain flexural stress vs. deflection curves. From these curves, flexural
strength (the maximum flexural stress attained) and flexural toughness (the area under the
load vs. deflection curve up to 20 % of the peak load in the post-peak) were determined.

Interlaminar bond strength tests were closed-loop interlaminar tensile strength

tests. Figure 4 shows a schematic of the ILB test employed. A 25.4 mm x 25.4 mm x 8
mm specimen was glued to the testing machine platens with an epoxy. Then a small
compressive load, 44.48 N, was applied to the platens until the epoxy cured. Next, a
uniformly increasing tensile load was exerted on the specimen at a rate of 0.004 kN/sec
until interlaminar (tensile) failure occurred. The ILB strength was taken as the maximum
force withstood divided by the cross-sectional area of the specimen.

Microstructural Properties
Moisture absorption capacity, before freeze-thaw cycling, and porosity, both
before and after freeze-thaw, \Vere evaluated to determine the microstructure of the
FRCB.

To determine moisture absorption, the dry weight of 175 mm x 50 mm x 8 mm

specimens was first determined. Then the specimens were immersed in water for four
days to ensure that a constant mass was reached and that all pores were filled. The
saturated weight of the FRCB was then measured. The moisture absorption was attained
by taking the difference between the saturated weight and the unsaturated weight,
dividing by the unsaturated weight and multiplying it by 100. Three replications were
made for each material.

Porosity measurements were determined by mercury intrusion porosimetry

(MIP) conducted on a Quantachrome Autoscan-33 Porosimeter. Two to four specimens
were tested for each material. After drying, the FRCB was placed in the testing chamber,
and the chamber vacuumed. Mercury was then introduced at increasing pressures. The
amount of mercury that intruded at each pressure interval was recorded by the computer
and written to a data file. MIP results indicate the total volume of pores and the
distribution of the pores. Here, the amount pores in the macropores range (pore radius >
I micron) is reported, since pores in this range are most responsible for freeze-thaw
damage.

RESULTS AND ANALYSIS

Before Freeze-Thaw Cycling

Flexural Perfom1ance -- Figure 5 presents typical flexural strength versus

del1ection curves for the three materials tested. Figures 6 and 7 show the average flexural
strength and toughness values, respectively. Before freeze-thaw cycling, Material A's
FRCB has the highest flexural strength and toughness. Manufacturer B's and C's FRCB
have similar 11exural strengths, and the toughness of Manufacturer B's FRCB is the

Copyright American Concrete Institute

150 Kuder and Shah

lowest. These results indicate that pressure alone does not control the flexural strength or
toughness of the FRCB before freeze-thaw cycling.

ILB -- Figure 8 presents the ILB for the three materials. Material C, the
unpressed material, has the lowest ILB, suggesting that pressure might improve ILB.
However, it is important to note that there is a significant difference in the ILB strengths
of the two pressed FRCB. Manufacturer A's ILB is 300 kPa larger than Manufacturer
B's ILB.

Moisture Absorption -- In Figure 9 the moisture absorption is shown. All FRCB

absorb close to 30% of their weight in water, indicating that pressure does not influence
moisture absorption. Regardless ofthe manufacturer, all FRCB consists of large amounts
of cellulose fibers and have a laminated structure, making the comparable absorption
capacities seem reasonable. From this high absorption value, serious deterioration may
be anticipated, since a significant amount of water will be present during freezing and
thawing.

Porosity -- Porosity measurements indicate that pressure treatments alone do not

control the porosity of the FRCB. Manufacturer A's FRCB appears to be the least
porous, but Manufacturer B's is the most porous. Figure 10 shows the total porosity of
each material, given as the cumulative pore volume. Manufacturer A's FRCB has the
lowest porosity, while Manufacturer B's has the highest. In Figure II the macroporosity
of each material is given. Manufacturer A's FRCB has the lowest percent of macropores,
Manufacturer B's has the lowest and Manufacturer C's falls between the two.

After Freeze-Thaw Cycling

Flexural Performance -- Figure 12 presents the flexural strength of the FRCB

after freeze-thaw cycling. After only 50 cycles, Manufacturer C, the unpressed product,
has lost over 70% of its initial strength. The pressed products also lose a considerable
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

amount of strength after only 50 cycles, but they are still more than 50% stronger than the
unpressed product. These results suggest the pressure treatments do improve flexural
performance after freeze-thaw cycling.

ILB Strength -- As can be seen in Figure 13, the ILB of all materials is severely
reduced after only 50 freeze-thaw cycles, demonstrating the susceptibility of the FRCB to
freeze-thaw damage due to its laminated structure. Both of the pressed products have
fairly constant and similar ILB strengths from 50 cycles and on. However, after 50
cycles, the ILB of Material C can no longer be measured because it is delaminating on its
own.

Visual observations during the three-point bend test also confirm the severe
reduction in ILB for the unpressed FRCB. Figure 14 clearly shows the interlaminar shear
failure of Manufacturer C's FRCB after 200 freeze-thaw cycles, while delamination is
not observed for the pressed material. Up to 300 freeze-thaw cycles, such a failure was
not observed for Manufacturer A's or B's FRCB. These results suggest that ILB is

Copyright American Concrete Institute

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
Thin Reinforced Cement-Based Products 151
improved by pressure after freeze-thaw cycling, although it may be the interlaminar shear
strength that is significant. Some preliminary tests were run to determine the interlaminar
shear properties of the FRCB by a short-span three-point bend test. However, due to the
small thickness, interlaminar shear failure could not be achieved.

Porosity -- Figures 15 and 16 present the cumulative pore volume and

macroporosity, respectively, for each FRCB, both before and after freeze-thaw cycling.
Opposing trends are observed. Before freeze-thaw Manufacturer A's FRCB was the least
porous, however, after freeze-thaw it is the most porous. Conversely, Manufacturer B's
material went from being the most porous to the least porous, both in terms of total
porosity and macroporosity. The unpressed material, produced by Manufacturer B, is
still in between the two pressed materials.

Changes in porosity could be due to an increase in interlaminar space, the

distance between fibers and the matrix or the overall porosity. Manufacturer A's FRCB
gains the most porosity, while Manufacturer B's material shows the smallest change in
porosity. The porosity measurements indicate that pressure treatments do not control the
porosity of FRCB after freeze-thaw cycling. In addition, porosity alone is not an
indicator of freeze-thaw durability, since the unpressed material, Maunfacturer C's
FRCB, is less porous than Manufacturer A's, while having a lower flexural strength and
an immeasurable ILB.

DISCUSSION

Previous work conducted by the authors showed that pressing FRCB improved
the mechanical properties before freeze-thaw testing when commercially-produced FRCB
made by the same manufacturer was pressed at a range of pressures (19). Flexural
strength and toughness increased due to an improvement in the fiber-matrix bond and the
possible densification of the material. In addition, ILB strength was significantly higher
and porosity was decreased. Such a clear improvement in properties is not seen here,
probably due to differences in processing.

After freeze-thaw conditioning, the results indicated that flexural strength was
not improved by pressure, but that 1LB strength was improved. ln addition, it was shown
that porosity was not affected by pressure. In this commercial comparison, an
improvement in ILB due to pressure after freeze-thaw cycling is seen, as is the lack of
improvement in porosity. However, flexural strength did appear to be improved in the
present work from the pressure treatments. Again, this highlights the difficulty in
comparing commercially-produced materials due to differences in processing parameters.
The magnitude and method of pressure applications are unknown as are the exact details
of the board formation and composition.

SUMMARY AND CONCLUSIONS

The research conducted demonstrates the weakness of FRCB to freeze-thaw

conditions and the role of the laminated structure in the deterioration of the material.

Copyright American Concrete Institute

152 Kuder and Shah

After only 50 freeze-thaw cycles, the majority of all the FRCB's ILB strength is lost and
a severe reduction in flexural strength is seen.

In addition, a number of conclusions may be drawn about the effects of pressure on

the durability of commercially-available FRCB:

1. Before freeze-thaw cycling flexural performance does not appear to be affected.

by pressure treatments. However, after freeze-thaw conditioning, pressure does
seem to improve flexural strength.

2. ILB tensile strength does not appear to be affected by pressure before freeze-
thaw cycling, but it is after freeze-thaw cycling. The unpressed material
delaminated after 100 freeze-thaw cycles, whereas the pressed materials
remained intact up to 300 cycles.

3. Pressure treatments may improve FRCB's interlaminar shear strength, which

may be an important material property of the FRCB, especially in defining
resistance to delamination due to freeze-thaw conditioning. However,
determining this parameter is difficult given its small thickness.

4. The effects of pressure on the porosity of FRCB is not clear, probably due to the
differences in board composition and manufacturing processes.

5. Analyzing the effects of pressure on commercially-available FRCB is difficult

due to differing production methods and materials.

It is important to note that the material tested here was subjected to extreme
conditions. The FRCB was both unsealed and unpainted. In the field, these two layers
would be present to protect the FRCB from the ingress of water. If water cannot get into
the material, deterioration will not occur. Once these layers become damaged, due to
handling or construction, freeze-thaw damage will occur. In addition, the material was
completely saturated throughout testing, whereas in the field saturation may not occur
and, even if it did, the material would eventually dry out.

Manufacturers have started to press FRCB in an attempt to minimize unsightly

delaminations caused by freeze-thaw cycling in the field, which would cause consumers
to lose confidence in the new construction material. The results from this work show that
the pressed FRCB does not delaminate on its own after 300 freeze-thaw cycles, while the
unpressed FRCB delaminates after only I 00 freeze-thaw cycles. Therefore, the primary
objective of pressing appears to be achieved. However, due to differences in board
composition and manufacturing processes, the exact benefits of pressing are difficult to
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

determine.

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 153

ACKNOWLEDGEMENTS

The authors gratefully acknowledge the financial support of CertainTeed

Corporation and the assistance of Bill Bezubic, Claude Brown and David O'Callaghan.
In addition, funding was also provided by NSF PATH grant #CMS-0122045.

REFERENCES
111
I. Kurpiel, F., "Rapid Growth of Cement-Cellulose Fiberboard (CFB)." 6
International Inorganic-Bonded Wood and Fiber Composite Materials Conference
( 1998): 55-60.
2. Kurpiel, F., "Diffusion of Cellulose Fiber-Cement Siding and Roofing into North
America." 5111 International Inorganic-Bonded Wood and Fiber Composite Materials
Co'!ference ( 1996).
3. Balaguru, P. and Shah, S.P. Fiber-Reinforced Cement Composites. McGraw-Hill,
Inc., New York, USA, 1992.
4. Bentur. A. "Fiber-Reinforced Cementitious Materials." Materials Science of
Concrete (1989): 223-283.
5. Akers, S.A.S., Crawford, D., Schultes, K. and Gerneka, D.A. "Micromechanical
Studies of Fresh and Weathered Fibre Cement Composites. Part 1: Dry Testing."
International Journal of Cement Composites and Lightweight Concrete II (2) (May
1989): 117-124.
6. Bentur, A. and Akers, S.A.S. 'The Microstructure and Aging of Cellulose Fibre
Reinforced Cement Composites Cured in a Normal Environment." International
Journal a_{ Cement Composites and Lightweight Concrete II (2) ( 1989): 99-107.
7. Bentur, A. and Akers, S.A.S. "The Microstructure and Aging of Cellulose Fibre
Reinforced Autoclaved Cement Composites." International Journal of Cement
Composites and Lightweight Concrete II (2) ( 1989): 111-115.
8. Pirie, B.J., Glasser, F.P., Schmitt-Henco, C. and Akers, S.A.S. "Durability Studies
and Characterization of the Matrix and Fibre-Cement Interface of Asbestos-Free and
Fibre-Cement Products." Cement and Concrete Composites 12 ( 1990): 233-244.
9. Soroushian, P., Shah, Z. and Marikunte, S. "Use of Kraft and Recycled Fibers in
Fiber-Cement Products." 3"1 International Inorganic-Bonded Wood and Fiber
Composite Materials Conference (1993): 9-19.
10. Tait, R.B. and Akers, S.A.S. "Micromechanical Studies of Fresh and Weathered
Fibre Cement Composites. Part 2: Wet Testing." International Journal of Cement
Composites and Lightweight Concrete II (2) (May 1989).
II. Soroushian, P., Marikunte, S. and Won, J. "Wood Fiber-reinforced Cement
Composites Under Wetting-Drying and Freezing-Thawing Cycles." Journal of
Materials in Civil Engineering 6 ( 1994): 595-611.
12. Nakamura, M., Fukushima, T. and Kamitani, M., "Microstructure and Frost
Durability ofCementitious Building Materials Reinforced with Non-Asbestos
Fibers." Journal of the Ceramic Society ofJapan I 00( 6) ( 1992): 858-863.
13. Venia, G. "Freeze-Thaw Perfonnance of Glass and Cellulosic Fiber-reinforced
Cementitious Boards." Proceedings Third International Conference on Concrete
Under Severe Conditions (2001 ): 522-529.
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

154 Kuder and Shah

14. Coutts, R.S.P. and Knightly, P., "Bonding in Wood Fibre-Cement Composites."
Journal of Materials Science 19 (1984): 3355-3359.
15. Akers, S.A.S., Garrett, G.G. "The Influence of Processing Parameters on the
Strength and Toughness of Asbestos Cement Composites." International Journal of
Cement Composites and Lightweight Concrete 8(2) ( 1986): 93-100.
16. Gupta, N.C., Paramasivam, P. and Lee, S.L. "Mechanical Properties ofCoir
Reinforced Cement Paste Composites." Housing Science 2(6) (1978): 391-406.
17. Coutts, R.S.P. and Warden, P.G., "Effect of Compaction on the Properties of Air-
Cured Wood Fibre Reinforced Cement." Cement and Concrete Composites 12
(1990): 151-156.
18. ASTM C 1185-99. "Standard Test Method for Sampling and Testing Non-Asbestos
Fiber-Cement Flat Sheet, Roofing and Siding Shingles, and Clapboards." American
Society of Testing and Materials, Philadelphia, PA, USA 2000.
19. Kuder, K.G. and Shah, S.P., "The Effects of Pressure on the Freeze-Thaw Durability
of Fiber-Reinforced Cement Board." Accepted for publication in A CI Materials
Journal, 2003.

List of Notations

FRCB- Fiber-reinforced Cement Board

ILB- Interlaminar Bond

MIP- Mercury Intrusion Porosimetry

Cement
38%

Silica'
53%
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Figure 1: FRCB Dry Composition by Weight

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 155

Bins containing dilute slurry

Figure 2: Hatschek Formation Machine

La~·cr ~ cvmpns~d
(tfmonol.a'\:"

Figure 3: Laminated Structure of FRCB

• •
FRCB Specimen

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Before Failure After Failure

Figure 4: Schematic of ILB Test

Copyright American Concrete Institute

156 Kuder and Shah

Deflection (mm)

Figure 5: Flexural Performance Before Freeze-Thaw

Figure 6: Flexural Strength Before Freeze-Thaw

~~------------·--------------------------,

.
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Figure 7: Flexural Toughness Before Freeze-Thaw

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 157

~r-------------------------------------~

Manufacturer A

~
~ 1500

ic
..
1:
ID
1000

::!
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

SOD

Figure 8: ILB Strength Before Freeze-Thaw

40 .---------

Figure 9: Moisture Absorption Before Freeze-Thaw

0.3 ... -------------~--------~~---------- ----

Manufacturer B

Figure 10: Cumulative Pore Volume Before Freeze-Thaw

Copyright American Concrete Institute

158 Kuder and Shah

E
0
lO

."e2 "
0
..e "
:J
0
>,

Figure 11: Macroporosity Before Freeze-Thaw

...
Freeze·thaw cycles
... ... ... ...
Figure 12: Flexural Strength After Freeze-Thaw
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

·~··················································································

....
·······························•···················•
Manufacturer A Manufacturer B Manuf.ctul1tr C
....
~ ....
-.,.

r:
5

= ...
= ...
'"
Figure 13: ILB Strength After Freeze-Thaw

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 159

Figure 14: FRCB During Three-Point Bend Test After 200 Freeze-Thaw Cycles
(a) Manufacturer A (b) Manufacturer C

Manufacturer A

.,
§ •...
0
> •oc1otes
f 8TOOCycln:$.
g_ ~-;
..,"'
.s
e""
"
t,)

Figure 15: Cumulative Pore Volume After Freeze-Thaw

"' A
B

•OCydi>•
•100Cydos
.§"'
;;
>
,.

Figure 16: Macroporosity After Freeze-Thaw

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

160 Kuder and Shah

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

SP-224-12

Crack Growth Resistance of Thin Mortar

Layers with Hybrid Fiber Reinforcement

by L. Sorelli. N. Banthia. and G. A. Plizzari

Synopsis: Hybrid fiber reinforcement of cement composites is rapidly emerging as an

innovative and prom ising way of improving mechanical performance and durability of
cement-based materials.

In the present paper, fracture behavior of medium, high and very high strength mortars
reinforced with hybrid fibers was experimentally studied by using contoured double
cantilever beam specimens. Different combinations of small steel fibers and fibrillated
polypropylene micro-fibers are investigated. These composites are very suitable for thin
sheet products such as roofing sheets, tiles, curtain walls, cladding panels, permanent
fonns, etc.

Aim of the paper was to study the influence of matrix strength, fiber type and fiber
combinations on the fracture toughness of the resulting fiber reinforced mortars.

Results indicate that some combinations of fibers and matrix strengths exhibit a higher
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

resistance to crack growth and evidence the contribution of polypropylene fibers to

mortar toughness.

Keywords: cement; fiber; reinforcement

161
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=University of Texas Revised Sub Account/5620001114, User=erur, ert
No reproduction or networking permitted without license from IHS Not for Resale, 01/26/2015 02:27:05 MST
Daneshlink.com
daneshlink.com

162 Sorelli et al.

Luca Sorelli received his PhD at University of Brescia in
2003, Italy. His current research consist of hybrid fiber
reinforced concrete, structural applications and finite
element modeling of high performance cementitious
composite.

Nemkumar Banthia, FACI, is a Professor of Civil Engineering

at the University of British Columbia (Canada). He is a
member of several ACI committees and chairs the ACI
Committee 544 on Fiber Reinforced Concrete. His primary
research interests consist of cement-based materials, fiber
reinforced concrete and fiber reinforced polymers,
shotcrete, strain-rate effects and impact, use of fiber
reinforced plastics in repairs.

Giovanni A. Plizzari, ACI Member, is a Professor of

Structural Engineering at the University of Bergamo
(Italy). He is a member of the FIB TG 4.5 "Bond Models" and
the RILEM Committee "Hybrid Fiber Concrete". His research
interests include material properties and structural
applications of High Performance Concrete.

INTRODUCTION
In the new breed of high performance cement based
materials, there has been great interest lately in the
development of Hybrid Fiber Reinforced Cementitious
Composites (HyFRCC) that combine different types of fibers
in a cementitious matrix [1]. The aim is to take
simultaneous advantages from the material properties of
each fiber type (multi-functionality) and from their
interaction (synergy) to optimize the mechanical and
physical performances of the composite [2-5].
A promising hybrid system of fibers concerns a combination
of steel fibers and polypropylene fibers. The former are
used to enhance strength and toughness properties [6] such
as flexural (modulus of rupture), shear [7), impact [8) and
fatigue strength [9]. The latter are commonly used to
reduce shrinkage cracking [10,11] and permeability [12] of
concrete; in fact, bundles of fibrillated polypropylene
fibers open during concrete mixing and separate into
millions of multistrand filaments that are able to mitigate
crack formation due to plastic shrinkage. Vondran and
Webster [12] found that a volume fraction (Vf) of 0.2% of

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 163

polypropylene fibers markedly reduce both the permeability
and the plastic shrinkage cracking.
Fibers may influence the fracture mechanism in a concrete
structure [3,13]. In fact, small-diameter fibers, here
defined as micro-fibers, may delay the fracture process by
which the micro-cracks coalescence to form large
macroscopic cracks [3,14]. Furthermore, micro-fibers modify

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
the crack pattern by transforming the macro cracks into a
network of smaller and narrower cracks.
A combination of small synthetic (polypropylene) fibers and
steel fibers could be used to yield a hybrid system that
may prove to be an interesting material for thin concrete
or mortar overlays for structural repair and retrofitting
[15]. The enhanced toughness, the reduced plastic shrinkage
cracking and the lower water-permeability could be highly
advantageous in producing a durable thin repair or product.
The use of short fibers in substitution of conventional
reinforcement (reinforcing bars or welded mesh) may allow a
reduction of labor costs.
In the present work, fracture behavior of thin mortar
layers with a combination of small steel and polypropylene
fibers is experimentally investigated by performing
Countered Double Cantilever Beam test (CDCB) [16].
Furthermore, the research aims to study the influence of
matrix strength on the mechanical behavior of concrete with
hybrid fibers.
The chosen amount of polypropylene fibers was higher than
the amount commonly used for controlling plastic shrinkage
cracking (Vf=0.1-0.2%), with the aim of improving the mortar
toughness. A small thickness of the specimens was adopted
to better reproduce the fiber distribution in thin
cementitious elements.
Specimens with a relatively large size were tested to
reduce the size effects and to allow for a simpler
determination of the mortar toughness.
In order to better understand the fracture behavior and to
determine the constitutive laws for the materials adopted,
the experiments were simulated by Finite Element analyses
based on Non Linear Fracture Mechanics (NLFM) [ 17] .

MATERIALS

The mix compositions include 854, 980 and 1019 kg/m 3 of

cement ASTM Type I for Medium (MSM), High (HSM) and
Very High Strength Mortars (VHSM) respectively. The water-
cement-sand (with a maximum diameter size of 5 rnrn)
proportions, as well as the air entrainer and the

Copyright American Concrete Institute

164 Sorelli et al.

plasticizer contents of the three different types of
mortars are reported in Table 1.
Steel fibers (SF) and fibrillated polypropylene micro-
fibers (PP) were combined as reported in Table 2. A
reference concrete without fibers (MSMO, HSMO and VHSMO)
was also made. Properties of the adopted fibers are
reported in Table 3. The steel fibers have a circular cross
section and a straight shape. They are made from high
carbon steel and are coated with brass for corrosion
protection.
The cylindrical 100 mm, h = 200 mm) compressive
strength (fcl determined after 28 days of curing is reported
in Figure 1; notice that the average compressive strength
for the MSM was approximately 60 MPa, for the HSM was
approximately 95 MPa and for the VHSM was approximately 115
MPa.

SPECIMEN DESCRIPTION AND TEST SET-UP

Figure 2a shows a schematic of the Contoured Double

Cantilever Beam specimen that was adopted for the
characterization of the crack growth resistance. According
to LEFM assumption and a model based on the crack
equivalent, the CDCB specimen is shaped in such a way that,
by using Linear Elastic Fracture Mechanics (LEFM), the

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
Stress Intensity Factor is independent of the crack length
and the specimen allows for a stable crack propagation
under constant load [18, 19]. The CDCB specimen also leads
to more reliable compliance measurements since the
displacements are large and the critical loads are small
compared with tests on other types of specimen [ 18] . A
groove reduced the thickness of the middle section from 40
to 15 mm to better control the crack path (Figure 2).
Four CDCB specimens were prepared for each material. The
direction of casting was perpendicular to the surface of
the double cantilever beam specimen and the fresh mortar
matrix was poured while the mould was externally vibrated.
The load was applied vertically by the hydraulic jack of
the Instron machine with a stroke rate of 0.1 mm/min on a
steel wedge placed between two rollers at the top of the
specimen (Figure 3a). The Splitting Load (SL) is the
horizontal components of the total load [19] (Figure 3b).
In order to limit the vertical component of the applied
load that may influence the fracture
behavior of the specimen, the angle of the wedge was chosen
equal to 15° [20]. The coefficient of friction between
the wedge and the rollers was ignored since the wedge

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 165

surfaces were carefully machined (Figure 4). The Crack
Mouth Opening displacement (CMOD) was measured by means of
a resistive displacement transducer (clip-gauge) which was
fixed at the level of the loading points (Figure 3a) .
The applied load and the CMOD data were acquired using a
digital data acquisition system running at 5 Hz.

RESULTS AND DISCUSSION

The Splitting Load vs. CMOD curves are given in Figure 4

for MSM, in Figure 5 for HSM and in Figure 6 for VHSM
mortars. The plotted curves represent the average curves of
four specimens tested and were obtained by means of the
full least-squares fit Loess procedure (a locally weighted
regression smoothing algorithm).
Notice a significant toughness increase in medium strength
mortars (MSM) due to the presence of steel fibers (along
with a higher residual strength and a more stable behavior
during fracture) . Furthermore, the marked differences in
the shapes of curves obtained for the steel fibers as
opposed to those obtained for the polypropylene fibers can
be observed (Figure 4). In the latter case, the curves are
characterized by a lower peak load followed by a steeper
post-peak branch. However, the general enhancement in the
performance of the polypropylene fibers when added to steel
fibers in hybrid materials should be noted.
The same trend is confirmed for the High and Very High
Strength Mortars (Figures 5 and 6) . It can be observed that
the peak loads increase with the matrix strength,
especially for the steel fiber reinforced mortars.
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

The hybrid materials show higher toughness than the mortars

with 0.5% of steel fibers, but lower toughness than the
mortars with 1% of steel fibers. This is in substantial
agreement with other researches carried out on same types
of fibers under bending [5, 21].
While the steel fibers were pulled out from the three
different matrices, part of the polymeric fibers broke
during the fracture. However, the presence of the secondary
polymeric fibers enhanced the fracture energy (GF; defined
as the area under the load-CMOD curves divided by the
projected cracked area) of about 50% for all the
cementitious matrices considered (Figure 7).
This shows that polypropylene fibers in the matrix with
steel fibers allow for appreciable advantages in term of
toughness beside the expected reduction of shrinkage
cracking (synergy).

Copyright American Concrete Institute

166 Sorelli et al.

The significant increase of peak load in specimens of HSM
as compared to specimens of MSM can be observed. On the
contrary, no major differences are observed when the matrix
strength further increases; this is probably due to the
bond strength between steel fibers and the mortar matrix
that did not increase in the VHSM (with respect to the
HSM).
Table 4 reports the number of steel fibers counted in the
cross section. In the same table it is also indicated the
expected number of steel fibers assuming either an uniform
3D distribution or a 2D distribution according to [22]:

(l)

where N is the expected number of fibers bridging the cross

section, Vf is the volume fraction of fibers, Ac is the
concrete cross section, Af is the fiber cross section area,
is a constant that varies from 0.5 for a 3D distribution
to 0.64 for a 2D distribution [22]. The results show that
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

the number of fibers counted in the cracked section is

always closer to a 3D distribution.

MODELING

The experiments were numerically simulated by using a 2D

Finite Element model based on Non Linear Fracture Mechanics
(NLFM) to better comprehend the test results and to
identify the fracture parameters of the materials used in
the tests. A discrete crack approach based on the
fictitious crack model was adopted [17).
The Finite Element analyses were performed by using Merlin
[23] that considers the structure as many linear elastic
subdomains linked by interface elements that simulate the
cracks, whose position must be known a priori.
Interface elements initially connect the sub-domains (as
rigid links) and start activating (i.e. cracks start
opening) when the normal tensile stress at the interface
reaches the tensile strength (fctlof the material.
Afterwards, the crack propagates and cohesive stresses are
transmitted between the crack faces according to a stress-
crack opening ( -w) law (Figure 8) which is given as input
for the interface elements.
The CDCB was modeled by adopting 3148 three node triangular
elements (plane stress) for the elastic sub-domains (having
a thickness of 40 mm), linked by means of 67 interface
elements (having a thickness of 15 mm) . By assuming a 2D

Copyright American Concrete Institute

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
Thin Reinforced Cement-Based Products 167
model, the stress concentration present at the groove tip
was neglected (Figure 9) .
Figure 10 shows the mesh adopted for the CDCB specimens.
The stress-crack opening displacement relationships -w)
were approximated with bilinear laws herein (Figure 8). The
tensile strength of this law ( fct) was determined from the
experimental compressive strength according to the CEB
Model Code 90 [24). The experimentally determined fracture
energies (GF) (Figure 7) were used as input data. The other
parameters, namely the stress at the knee point ( 1 ) , the
crack opening at the knee point (w 1 ) were identified by an
inverse analysis based on the best fitting procedure [25).
Eventually, critical crack opening (Wcr> was determined.
Figure 11 shows a typical comparison between the numerical
and the experimental curves for the steel fibers (Vt=1%) in
the High Strength Mortar. The same figure exhibits the
deformed mesh at different loading stages as well as the
distribution of cohesive stresses over the ligament length.
It should be noticed that the crack tip opening
displacement at the peak load is around 0.14 mm and that
the fracture process zone involves most of the ligament
length. The large crack tip opening displacement explains
why, in the adopted specimens, the peak load is more
related to the fiber bridging mechanisms than to the matrix
strength.
The numerical and the experimental curves of the Splitting
Load versus the CMOD are plotted for all the MSM materials
in Figure 12; notice the excellent agreement between the
different curves. The same results are reported in Figures
13 and 14 for the HSM and VHSM mortars, respectively.
The best fitting parameters of the bilinear softening laws
as well as the modulus of elasticity are summarized in
Table 5.

CONCLUDING REMARKS

Splitting tests were carried out on Countered Double

Cantilever Beam specimens. Because of the cross section
thickness of 15 mm, these specimens seem suitable to
characterize the fracture behavior of thin concrete members
made of fiber reinforced concrete.
Experimental results indicated that steel fibers better
enhance the mortar toughness. However, the addition of
polypropylene fibers to a steel fiber reinforced mortar
increases the toughness of the composite for all matrix
strengths considered. In fact, the fracture energy (GF) of
the hybrid materials with 0.5% of steel fibers was improved

Copyright American Concrete Institute

168 Sorelli et al.

from 35% to 64% by polypropylene micro-fibers in the three
different matrices strengths (medium, high and very high) .
Considering the fact the polypropylene fibers better
control the cracks due to the plastic shrinkage, the
reduced permeability and the lower cost of polymeric
fibers, these hybrid composites seem very suitable for thin
concrete overlays for structural repair and retrofitting.
Furthermore, they can be conveniently adopted for thin
cementitious products, such as roofing sheets, tiles,
curtain walls, cladding panels, permanent forms, etc.
However, although synergy between the two fibers is already
apparent in the hybrids, further optimization attempts are
clearly warranted. Therefore, further research, which
considers plastic shrinkage permeability and thermal
effects, is necessary to optimize the combinations of these
fibers.
The fatigue resistance may also be improved by a hybrid
system where micro-fibers can be active as bridging
mechanism over the micro-cracks surrounding macro-fibers
and cause synergistic effects in the composite. In
addition, in case of a fire, when the free and chemically
bonded water is transformed in vapor, the polymeric fibers
will melt leaving canals through which water vapor can
escape from the boundary zones without spalling off the
concrete covers. This may guarantee the fire protection
required in structural applications.
Non Linear Fracture Mechanics is a satisfactory tool to
model the fracture behavior of these cementitious
composites where the fracture process zone involves most of
the ligament length of the specimen.

Acknowledgements

The authors would like to thank Mr. David Woomk for his
diligence and his enthusiasm in preparing the experimental
tests as well as the helpful support of the technicians of
University of British Columbia (Canada).
Thanks are also due to the Dow Chemical Company and the
Bekaert for supplying respectively the polypropylene and
the steel fibers.

REFERENCES

(l] Banthia, N., Yan, C. and Bindiganavile, V. 'Development

and application of high performance hybrid fiber reinforced
concrete' International Rilem Symposium on FRC, BEFIB 2000,
Lyon, September 2000, 471-480.
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 169

[2) Meda, A., Plizzari, G.A., Sorelli, L. and Banthia, N.
'Uni-axial and bending tests on hybrid fiber reinforced
concrete' Celebrating Concrete, International Symposia:
People and Practice, Dundee, 3-4 September 2003, 709-718.
[3) Lawler, J.S., Zampini, D., and Shah, S.P. 'Permeability
of cracked hybrid fiber-reinforced mortar under Load' ACI
Materials Journal, 94, 2002, 379-385.
[4) Markovic, I., Van Mier, J. G. M. and Walraven, J. C.
'Single fiber from the hybrid fiber reinforced matrices'
International Symposium on High Strength/High Performance
Concrete, 2, 2001, 1175-1186.
[ 5) Stroeven, P., Shui, Z., Qian, C. and Cheng, Y.
'Properties of carbon-steel and polypropylene-steel hybrid
fiber concrete in low-volume fraction range'
Proceedings 5th Conference of Recent Advances in Concrete
Technology, 2001, 713-731.
[6) Balugaru, P., and Shah, S.P. 'Fiber Reinforced Cement'
Composites, McGraw-Hill, 1992, 530.
[7) Casanova, P., Rossi, P. and Schaller, I. 'Can steel
fibers replace transverse reinforcement in reinforced
concrete beams?' ACI Materials Journal, 94 (5), 19971 341-
354.
[8) Shah, S. P. and Suaris, W. 'Strain-rate effects in
fibre reinforced concrete subjected to impact and impulsive
loading' Composites, 13, 1982, 153-159.
[9] Koermling, H.A., Reinhardt, H.W., and Shah, S.P.
'Static and fatigue properties of concrete beams reinforced
with bars and fibers' ACI Journal, 77(1), 1980, 36-43.
[10) Krenchel, H. and Shah, S. P., 'Restrained shrinkage
tests with polypropylene fiber reinforced concrete' Fibre
Reinforced Concrete Properties and Applications, ACI
Special Publication, 105, 1987, 141-158.
[11) Zollo, R. F. and Ilter, J. A. 'Plastic and drying
shrinkage in concrete containing collated fibrillated
polypropylene fibre' Developments in Fibre Reinforced
Cement and Concrete, Proceedings RILEM Symposium,
Sheffield, 198 6.
[12) Vondran, G. and Webster, T. 'Relationship of
polypropylene fiber reinforced concrete to permeability'
Permeability of Concrete, 1988, 85-98.
[13) Banthia, N., Bentur, A. and Mufti, A. 'Fiber
Reinforced Concrete: Present and the Future' Canadian
Society for Civil Engineering, Montreal, 1998, 64-97.
[14) Boulet, D., Pleau, R., Rougeau, P. and ~octet, R.
'Flexural behaviour of ultra high-performance cementitious
composites reinforced with different types of steel fibers'

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

170 Sorelli et al.

International Rilem Symposium on FRC, BEFIB 2000, Lyon,
September 2000, 759-768.
(15] Banthia, N. and Nandakumar N. 'Crack growth resistance
of Hybrid Fiber Reinforced Cementitious Composite' Cement
and Concrete Composites, 25, 2003, 3-9.
(16] Mostovoy, S., Crosley, P. B. and Ripling E. J. 'Use of
crack-line-loaded specimen for measuring plain-strain
fracture toughness' Journal of Materials, 2 (3), 1967,
661-681.
[17] Hillerborg, A., Modeer, M. and Petersson, P. E.
'Analysis of crack formation and crack growth in concrete
by means of fracture mechanics and finite elements' Cement
and Concrete Research, 6, 1976, 773-782.
[18] Visalvanich, K. and Naaman, A.E. 'Fracture methods in
cement composites' Journal of the Engineering Mechanics,

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
ASCE, 107, 1981, 1155-1171.
[19] Banthia, N and Genois, I. 'Controlled crack growth
tests for optimization of micro-fiber reinforced cement
composites' ACI Special Publication, 201, 2000, 55-74.
[20] Brhtiwiler, E. and Wittmann, F.H. 'The wedge splitting
test, a Method of Performing Stable Fracture Mechanics
Tests' Engineering Fracture Mechanics, 35, 1990, 117-126.
[21] Bentur, A., Mindess, S. and Skalny, J. 'Reinforced of
normal and high strength concretes with fibrillated
polypropylene fibers' International Conference 'Fibre
Reinforced Cements and Concretes: Recent Developments', 18-
20 September, Cardiff, UK, 1989, 229-239.
[22] Wang, Y., Li, V. C. and Backer S. 'A statistical
tensile model of fibre reinforced cementitious composites'
Composites, 20(3), 1989, 265-274.
[23] Reich, R. W., Cervenka, J. and Saouma, V. E. 'Merlin,
a three-dimensional finite element program based on a
mixed-iterative solution strategy for problems in
elasticity, plasticity, and linear and non linear fracture
mechanics' EPRI, Palo Alto, CA, 1994;
http://civil.colorado.edu/-saouma/Merlin.
[24] CEB 'Model Code 90' CEB Bulletin, N° 213-214, 1993.
[25] Roelfstra P. E. and Wittmann F. H. 'Numerical method
to link strain softening with failure of concrete' Fracture
Toughness and Fracture Energy of Concrete, Amsterdam, 1986,
163-175.

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 171

Table 1 . Mix design of the three types of mortar .
Mortar Mixes W:C:Sand Plasticizer Silica Fume
[m1/kgceml (%]cern.
HSM 0.45 : 1 : 1 - -
VHSM 0.25 : 1 : 1 8.0 10%
UHSM 0.16 : 1 : 1 20.0 20%

Table 2. Fiber combinations and of the different types

-
of mo.,...tar
Mix Code Fiber SF Fiber PP
[%] [%]
MSM 0 - -
..,
.c MSM 1 0.5
.c "" MSM 2 1.0 -
..."""..," MSM 3 - 0.5
"' " MSM 4 - 1.0
" MSM 5 0.5 0.5
.c HSM 0 - -
...O>.C.., HSM 1 0.5 -
"' ""' HSM 2 1.0 -
"'"
..,.., HSM 3 - 0.5
"" HSM 4 - 1.0
""'"'
"' HSM 5 0.5 0.5
VHSM 0 - -
""' .c..,
w
.... .c "'
VHSM
VHSM
1
2
0.5
1.0
-
-
.... "
::> "" " -
...>. = ..,...
VHSM 3 0.5
U) VHSM 4 - 1.0
"
> VHSM 5 0.5 0.5

Table 3. Geometrical and mechanical characteristics of

the steel and polypropylene
- fibers.
Fibers Fiber <I> Lf Lfl<i>f fft EF p
Code [J.Illl] [mm] [MPa) [GPa) [kg/m 3 ]
Steel SF 400.0 19.0 45.0 2000 210 7850
Polyprop. pp 30.7 12.5 407 375 3.5 900

Table 4. Average r.umber of steel fibers intercepting the

crack section with the relative standard deviation
t<'-aterial Steel 0.5% Steel 1.0% Hybrid
MSM 61.75 (±36%) 102.8 (±34%) 61.7 (±29%)
HSM 54.0 (±21%) 148.0 (±16%) 48.8 (±17%)
VHSM 61.0 (±12%) 117 .o (±7%) 54.3 (±10%)
Expected fibers
2D-distribution 91 182 91
Expected fibers
3D-distribution 71 142 71

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

172 Sorelli et al.

Table 5. Constitutive parameters for the materials
adopted.

Materials E, foe w, 0] Wor

[GPa] [MPa] [mm] [Mpa] [mm]
MSMO 11 3.64 0.015 0.350 0.150
MSM1 17 4.75 0.014 0.731 5.500
MSM2 17 4.16 0.011 1. 366 5.000
MSM3 17 4.16 a. o1s 0.274 3.000
MSM4 17 3.84 0. 015 0.401 3. 800
MSMS 17 4.41 0. 013 0.934 6.300
HSMO 19 5. 71 0.010 0.35 0.15
HSM1 19 6.60 0.009 0.73 4. 50
HSM2 19 5.75 0.007 2.17 5.00
HSM3 19 6.35 0. 010 0.27 3.50
HSM4 19 5. 71 0.010 0.40 3.50
HSMS 19 5.61 0. 009 1. 07 5.00
VSMO 21 7.21 0.012 0.80 0.22
VSM1 21 6.70 0.005 1. 00 4. 50
VSM2 21 6.96 0.007 2.20 3.70
VSM3 21 6.54 0.012 0.30 3.00
VSM4 21 6.60 0.012 0. 68 2.40
VSMS 21 6.40 0.001 1. 47 4.20

&:' 1400
:!
£
t:l>
c
~
~ 800
"gj
~
~
20.0

00 VHSM

pol)'1%
Hyb.W

Figure 1. Compressive strengths for different materials adopted.

1·1.

(~ (bl
Figure 2. Schematic of a Countered Double Cantilever Beam specimen (a); schematic
crack path (b).
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 173

(a) (bl

Figure 3. A CDCB specimen under loading (a); load transmitted by the steel wedge (b).

CMOO[mm]

Figure 4. Splitting Load vs. CMOD curves experimentally determined from MSM fiber
reinforced mortars.

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

174 Sorelli et al.

0 2 4 e
CMOO!mmj

Figure 5. Splitting Load vs. CMOD curves experimentally determined from HSM fiber
reinforced mortars.

0.5%-- • 0.5%

()
0 2 4 6 8
CMOO[mm]

Figure 6. Splitting Load vs. CMOD curves experimentally determined from VHSM fiber
reinforced mortars.

OMSM(OO~)
• HSM {95 Mpa)
Ill! 1/HSM 1115 Mpo)

~
o• .. ~ ~~~..
Plain mort3r s1eal 0 5% otae"'1'4
. r•.. . . . . . l..l
poly 0.5% poly 1% hyb!id

Figure 7. Fracture energy Gr values for materials adopted.

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

f.,

,
I
;~-,_E.
I
I
+
(jb
Thin Reinforced Cement-Based Products

f.,

<>•
-- ·-----~-----)>
175

6 w, Wcr W

Figure 8. Constitutive laws for discrete crack model.

Figure 9. 3D stress distribution due to the groove.

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Figure I 0. Mesh of the specimen and pre-imposed crack line.

Copyright American Concrete Institute

176 Sorelli et al.

4.C

1.C
- E>rperimootai CUfW$
·--·• Nwmofleaf wrve

0.0

CMOCl[mmj

Figure 11. Numerical and experimental curves in terms of Splitting Load and CMOD for
the Medium Strength Mortar with 1% of steel fibers.

4
CMOOjmm]

Figure 12. Experimental and numerical Splitting Load versus CMOD curves for MSM
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

mortars.

/~'-5~ -+0.5~
I

G
0
CMOOimm]
•
Figure 13. Experimental and numerical Splitting Load versus CMOD curves for HSM
mortars.

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 177

ESteelllber ! P~-·
- i ;;'::
--..- ......!.-·····-·--

. ·.··.<
~· .. ··,.

4
CMOO{mm)

Figure 14. Experimental and numerical Splitting Load versus CMOD curves forVHSM
mortars.

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

178 Sorelli et al.

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

SP-224-13

Modeling of Cement Based

Composite Laminates

by B. Mobasher

Synopsis: Techniques for modeling the mechanical response of thin section cement-
based composites intended for structural based applications are presented using a
micromechanical approach. A layer model is used and the property of each layer is
specified based on the fiber and matrix constituents in addition to the orientation and the
stacking sequence in each lamina. The overall axial and bending stiffness matrix is
obtained using an incremental approach which updates the material parameters. The
simulation is conducted by imposing an incremental strain distribution, and calculating

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
the stresses. A stress based failure criterion is used for the three failure modes of
initiation of cracking, ultimate strength of matrix, and ultimate strength oflamina. As the
cracking saturates the specimen, it results in a gradual degradation of stiffness. A
continuum damage model based on a scalar damage function is applied to account for the
distributed cracking. The model predicts the response of unidirectional, cross ply and
angle ply laminae under tensile loading in longitudinal and transverse directions. The
load-deformation responses under tension and flexure are studied. It is shown that by
proper selection of modeling approach, parameter measurement, and theoretical modeling,
a wide range of analysis tools and design guidelines for structural applications of FRC
materials are attainable.

Keywords: cement; cementitious composites; concrete; cracking; fibers

Copyright American Concrete Institute

179
Provided by IHS under license with ACI Licensee=University of Texas Revised Sub Account/5620001114, User=erur, ert
No reproduction or networking permitted without license from IHS Not for Resale, 01/26/2015 02:27:05 MST
Daneshlink.com
daneshlink.com

180 Mobasher
ACI member Barzin Mobasher, Ph.D. is an associate professor of civil and environmental
engineering at Arizona State University. He is a member of ACI Committee 544, Fiber
Reinforced concrete, 549 Thin Reinforced Products, and 446, Fracture Mechanics. His
research activities include fiber reinforced concrete, toughening mechanisms, and
modeling of durability.

Introduction

In order to commercially utilize new composite materials in civil engineering

applications, simple and effective analysis and design guides are needed. Theoretical
models are also needed to predict the response of laminated composites in order to better
understand the interaction between the various phases and aid in the design of the overall
structural system. The present work presents a general framework of analysis and design
for modeling the uniaxial and flexural response of composite laminates. This
methodology can be used for new composite materials or strengthening components of an
existing structure.

In the proposed theoretical approach, the degradation of stiffness is considered using a

strain based scalar damage-softening model. Three zones of behavior are considered for
the matrix phase, including the elastic range, the range of stiffness degradation due to
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

initiation and generation of parallel crack formations and the strain softening range. The
load carrying capacity of the matrix phase in each lamina decreases after cracking and the
stiffness of the lamina degrades such that the composite response asymptotically
approaches the levels predicted by the ply discount method. An exponential strain
softening response for matrix in the post cracking range is considered and for a lamina
with its matrix phase in unloading mode, a proportional unloading for the stresses in
other directions is assumed.

Equivalent Elastic Lamina Formulation

A general approach for the treatment of composites made with various fiber and matrix
materials as continuous and cross ply laminates is used. Each lamina is modeled as an
orthotropic sheet in plane stress with direction "1" representing the longitudinal direction
of alignment of fibers, and direction 2 representing the transverse direction as shown in
Figure I. Parameters hk and hk+l represent the coordinates and top and bottom of lamina
number "k" in a stack of "n" laminates. Angle El represents the orientation of fiber
direction with respect to the direction of application of load, hence a 0 degree lamina
represents the load being applied in a direction of the fibers, and 90 degree lamina
represents the load being applied transverse to the direction of the fibers. The fiber is
assumed to be linear elastic, and the effect of fiber volume fraction is incorporated in the
elastic properties of each lamina. Based on the layer model, the property of each layer is
specified using the material properties and volume fraction of components. Using the
stacking sequence the overall axial and bending stiffuess matrices are obtained. The
equivalent elastic stiffness of each lamina is obtained using the sum of the contributions
from each phase to the overall value. Depending on the state of strain (normal and shear)

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 181

and curvature distribution, strains at the top and bottom of the lamina are calculated. The
strain distribution is applied to the orthotropic model to calculate ply stress.

In the elastic range the rule of mixtures for longitudinal modulus and the Halpin-Tsai [ 1)
estimates oftransverse modulus is applicable. This zone is terminated by initial cracking
of the matrix phase using a stress-based criterion [2) at stress levels designated as au. lt
is furthermore assumed that the load carrying capacity of the matrix is not exhausted
completely and as microcracking in the composite takes place, the stiffness degrades
according to a single scalar damage parameter 'ro'. The form of the evolution of the
damage parameter as a function of strain is expressed as:

(1)

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
The form of the function in equation I was used based on a model proposed by Karihaloo
and Fu [3] is used to formulate the damage vs. strain relationship as shown in Figure 2.
This empirically based damage evolution approach is used in conjunction with a model
by Horii [4] and also Nemat Nasser and Hori [5] to estimate the degradation of stiffness
as a function of strain as shown in Equation 2. In this equation, the damage parameter Ql
is calculated at various strain levels with constants a, p, Hand ro 1 as shown in Equation
2. The values of these constants are of a= 0.16, p= 2.3, and ro 1 = £ 11 H = 0.05, where H
is the gage length of the specimen used. au and Eu = o 11 /Emo were used to represent the
ultimate strength, and strain at failure under uniaxial tension for the paste in an
unreinforced condition. Within the cracked matrix range, as the strain is increased, the
stiffness of the matrix decreases in terms of a damage evolution law as proposed by Horii
et al.[4]. The stiffness defined as a function of damage is Em(ro) and expressed m
equation 2 as a function ofuncracked matrix elastic modulus Emo:

E,(m) = 16 ~
(2)
1+ - m(l- v- )
3 m

This value is used in the rule of mixtures to obtain the longitudinal stiffness ofthe lamina
in the longitudinal direction E 1(ro), as defined in Equation 3. Calculation of the transverse
modulus E2 and v12 were achieved using the Halpin-Tsai equations as shown in Equation
s
2. The value of was set equal to 2 in the present study. This is because, the fabric used
is circular [6).

Copyright American Concrete Institute

182 Mobasher
E1(w) =E1 V1 + Em(w)(l- V1 ) (3)

(4)

The stress in the matrix phase beyond the elastic range is calculated incrementally as:

i
a: (w) = 0"11 + LEm(w)(en -&n_1)
n=l
(5)

Equation 5 computes the stress using an incremental approach of adding the products of
strain increments by the effective stiffness at that level. The degraded stiffuess at each
strain value up to a strain level defined as Emu are used. Based on this approach there is a
gradual decrease in the stiffness of the matrix beyond the plain matrix crt! until the
ultimate strength of matrix O"mu· This relationship is maintained until a damage level
defined by ro0 is reached at the ultimate strength of matrix in the presence of fibers. The
parameter Emu is obtained using the ACK approach [6] which predicts the strength of
matrix phase in the presence of fibers. In this approach y is the fracture toughness and r
is the fiber radius. In the current study, y = 0.5 N-mm has been used. This approach has
been verified to be applicable for the cement based materials as it has been clearly shown
that the strength of the matrix is increased in the presence of fibers. [7]

(6)

Beyond this level, the response is dominated by localization of the matrix phase, and is
referred to as the softening zone. The stress in the strain-softening zone asymptotically
approaches a level of zero, after which the model is comparable to the ply discount
method, which totally neglects the stress in a cracked layer. In this zone the matrix
cracks widen and while there may be no localization, the strain softening region is
defined as a zone where the response is governed by a smeared crack model. The stress
capacity is assumed to an exponentially decaying function of the maximum stress. The
choice of the exponent parameter affects the rate of drop of the stress as a function of
strain. This response is modeled as:

(7)

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 183

where "w" represents the exponent coefficient affecting the rate of decay in stress from
the peak composite stress. The range of values of w=S0-150 was used in the simulation
of the data in this study. Clearly the definition of strain in this region is gage length
dependent and the present approach uses the mean strain over the length of several cracks
in the matrix. As the specimen undergoes strain softening, an exponential decaying
stiffness similar to Eq. 7 utilizing the stiffness at peak was used. The modulus Em,
computed for each strain level E, was hence proportional to the reduction of the stress
from the peak value using Eq. 7.

Failure Criteria for Lamina

It is known that matrix in the 0 degree plies may be subjected to significant parallel
microcracking due to the bridging effect of fibers. The matrix phase in the 90 degree
plies loaded in tension may also be subjected to parallel cracking due to the shear lag of
adjacent layers. A cracked matrix in a degree ply may carry a significant amount of
stress due to fiber bridging, whereas a cracked matrix in a 90 degree layer may be stress
free due to Jack of fiber bridging. Therefore the initial cracking and final cracking of the
matrix must be differentiated. Additionally, the complete failure of lamina due to the
failure of the fiber phase must also be considered. For an off-axis lamina subjected to
shear, the matrix phase may fail in a brittle manner due to the formation of a single shear
crack. The failure criterion for the first cracking of matrix and final cracking of matrix
based on the state of stress and represented as the yield surface, F 1 and F2 :

cr 1 ~ crmu

After each incremental loading, stresses in the lamina were checked against the failure
surface to update the material properties for the subsequent iteration. The second yield
surface F2 was used to address the strength of the matrix in the presence of fibers or crmu·
For a unidirectional lamina subjected to tension, assuming that the matrix phase has
cracked significantly, the ultimate tensile strength was set equal to the strength of the
fiber phase, and represented as:

(10)

Generalized Load-Displacement for the Composite Response

The constitutive relations for a general orthotropic material require the compliance
matrix, S, or its inverse the stiffness matrix, Q, which relate the stress and strain within a
lamina loaded in its principal directions [8]. Since the present model updates the elastic

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

184 Mobasher
stiffness of the matrix due to cracking, an elastically equivalent compliance matrix S
was defined where the bar indicates use of updated elastic properties. In the term S;ik ,
parameter "i" represents the load increment, "j" the direction of applied strain, and "k"
the observed stress. The stress strain relationship was represented in incremental form
for each loading increment i, as:

a! =(s)k t 11&~ + a!-l

(II)
In matrix form:

l
0 11&1 0'1

0
s66
-lr 11yl21+ r 1
11&z

i
O'z

rl2 i-1
(12)

where,

1 1
Sz2 = E2(w) S66 - Gl2(w)

(13)

By inverting the compliance matrix, S, the stiffness matrix, Q is obtained which relates
the stresses to strains for each lamina loaded in principal material directions.

(14)

For a composite laminate consisting of several laminae, each with a fiber orientation
of(;m, where m represents the first to the n1h ply, classical lamination theory results in
derivation of laminate stiffness components as:

~j = L Qijm (hm - hm-1 ),

m::;J

The form of submatrices A , B and D is discussed by Agarwal and Broutman [2], where
A represents the extensional, f5 the bending, and B the coupling stiffnesses. With
knowledge of laminate strains and curvatures, the stress distribution per lamina is
computed for each loading step in an incremental fashion. M represents the moment per
unit length, N the force per unit length of cross section, s 0 and K represents the

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 185

midplane strains and the curvature of the section respectively. The strains and forces
were updated incrementally according to the matrix form representation:

(16)

For each iteration, the incremental loads and strains are determined and used to update
the previous increment values. The applied load in the x direction at the i1h interval of the
j 1h lamina was represented as N ix.i according to:

N 1 .=
X,/
N 1. +
X,/-]
~ 1 .= N;.
X,/ X,/-]
+[A] . [~c I
0
] (17)

Similarly,

M;X,/.= M 1.
X,l-]
+ ~M ; .= M X,l
1. +
X,l-]
[J5] . (~K] I
(18)

After the geometry of the laminate in terms of number of layers and their orientation is
defined, the solution algorithm imposed the strain and curvature distributions
incrementally. At each increment of the strain, the stiffness is calculated and used to
calculate the stress. The stress is checked against the failure criteria for plain matrix
failure, bridged matrix failure, and the composite failure. If the failure criteria were met,
then the stress level and the stiffness of that layer are adjusted according to the
constitutive response. Subsequent loading of a cracked layer results in a change in the
magnitude of the damage parameter. This indicates that at any stress level, the
degradation of elastic properties is primarily related to the magnitude of crack density
and overall strain response. Using the updated damage parameter, the quasi-elastic
- - -
stiffness parameters A, Band D are obtained and used to calculate the load and
moment for that increment. The procedure is repeated for the next strain increment. A
complete description and the parametric evaluation of the model are provided elsewhere.
[9]

Performance of Model: Simulation of Tensile Load

Several case studies involving various systems are presented to evaluate the
applicability of the model to composite materials under tension and bending. Figure 3a
and 3b present the simulated and experimental results for 0/90/0 and [0/45/-45/90/90] 5
stacked laminates subjected to a unifom1ly applied tensile strain level. A constant strain
level is imposed across the depth of the cross section. As seen in Figures 3 a and b, the
cracking starts with matrix cracks forming in the 0 degree and 90 degree layers. This is
followed by cracking in the ±45 degree layers due to shear. Damage is allowed to
accumulate in the 0 degree layers due to multiple matrix cracking in accordance to the
damage evolution law. The loading in the transverse direction (90 degree layers) is
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

186 Mobasher
limited to the ultimate tensile strength cr12. Note that as the fiber volume fraction is
increased, the response of specimens in carrying the forces and distributing the cracks
beyond the initial cracking phase are also enhanced. As the damage accumulation
increases, it results in a reduction in stiffness for the overall composite. The load
carrying capacity extends well beyond the matrix-cracking point and as damage
accumulates, stiffness decays. The stress in the longitudinal layers increases to a
maximum level determined by the fiber fracture strength, or an effective strength of V1
O'fu· Successive failure of 0, 90, and 45 degree layers is apparent in the angle ply samples
as shown in Figure 3.b. Note that in the transverse direction the stiffness and strength are
both significantly lower than the 0° layers. The stiffuess degradation due to damage
results in a non-linear response which is also shown in the load vs. deformation response;
however, this is not clearly visible due to the high relative stiffness of glass as compared
to the cement matrix.

Figure 4 represents a comparison of the theoretical predictions with experimental

results for continuous AR glass fiber systems for both unidirectional and cross ply
lamina. The experimental procedures are described in detail elsewhere [IO]. A uniform
strain is imposed in the principal material direction I across the I 8 mm thickness of the
cross section at several stages. As the ultimate strength of the matrix phase is reached,
there is a shift in the slope of the stress strain response, also known as the Bend Over
Point (BOP). The load carrying capacity extends well beyond the matrix-cracking phase
and as damage accumulates the stiffness decays. Results are also compared with a the
response of a [0/90], stacked lamina (V t=9% ). The loading in the 90° layers is limited to
the ultimate tensile strength a 12 • This results in a lower stress in the 90° layers. The
maximum load is attained when the stress in the remaining 0° longitudinal lamina reaches
a stress equal to the effective strength of the fiber phase or VfO'fu.

The model was further extended to composites with fibrillated polypropylene fibers. The
values of Em=30000 MPa , Et = 7000 MPa, Vm = 0.18 Vf = 0.25, and lamina strength of
atl = O't2 = 6 MPa were used. Figure 5 represents the model predictions for the response
of unidirectional (0), 0/90/0, and 90/0/90 laminates with polypropylene fiber composites.
There is a major drop in the stiffness of the composite as the strength of the matrix is
reached at the bend over point. This is attributed to the low stiffness of the
polypropylene fibers. As a 0 degree lamina is replaced by 90 degree layers, it is observed
that both the first crack strength and also the post BOP stiffness drop markedly; however,
the benefit of this lay up arrangement is found in improvements in transverse properties
of the layers. The response exhibited in 0/90/0, and 90/0/90 laminates demonstrates the
behavior of an ideal composite for use under a biaxial loading condition since both
transverse and longitudinal directions are ductile and strong; whereas, the 0 degree
laminates show a very strong and ductile response in the longitudinal direction, however,
the transverse response is brittle.

Figure 6 presents a comparison of model predictions with experiments for unidirectiona

polypropylene fiber composite systems [10]. Similar to the case of glass fabrics, at tht
fiber volume fraction of 6% pp fibers, a BOP strength level of 8 MPa is obtained. Due t<
the high ultimate strain capacity of the polypropylene fibers, it is observed that the overal
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 187

strain in the sample may be of the order of several percent. The choice of the crad
spacing-stiffness degradation model in the matrix is quite important in the response o
these composites. The results shown are for a constant strain softening coefficient o
w=50, and a=5, and P=0.8 used in Eq I.

Simulation of Flexural results

Simulation of the flexural load-deflection response of a unidirectional laminate is shown

in Figures 7 and 8. The various stages of loading are obtained by increasing the
magnitude of strain that changes linearly across the thickness of the specimen. The
longitudinal stress distribution results in cracking in the tension zone and is followed by
distributed cracking and strain softening. The compression zone is assumed to carry the
stresses in a linear way. In the present analysis the neutral axis is obtained by solving for
the equilibrium of internal forces. Using the location of the neutral axis and the strain at
the extreme fiber, the resulting moment-curvature response of the cross section can be
obtained by integrating the first moment of the stress distribution through the thickness,
while the curvature distribution is obtained from the strain magnitude.

Figure 7 represents the effect of fiber volume fraction on the flexural moment curvature
response of a unidirectional laminate. The response of a composite with 6% AR Glass
fibers shows three distinct levels of cracking due to the failure of each lamina in tension.
The moment curvature responses indicate the improved deformation capacity of
composites with higher fiber fractions. Note that as the fiber volume fraction increases,
the initial stiffness remains the same, however, the point of first cracking is increased.
Above a certain critical level of fibers, it is possible for the composite to carry loads
beyond the first cracking load or the proportional Elastic Limit (PEL).

The response of a unidirectional specimen is compared to a [0/90/90/0] composite in

Figure 8. The cross-ply laminated composite exhibits cracking and loss of load carrying
capacity. This leads to nonlinear behavior.

Conclusion

A theoretical model is presented to predict the response of composite laminates

subjected to axial loads. The model utilizes composite laminate theory subjected to
material degradation by means of a scalar damage parameter. Several case studies are
presented and theoretical results are compared to experimentally obtained data and
indicate a good agreement for several lamina configurations.

REFERENCES

Halpin, J.C., and Tsai, S. W., ( 1967) "Environmental Factors in Composite

Materials Design," Air Force Materials Research Lab., Technical Report,
AFML-TR-67-423.
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

188 Mobasher
2 Agarwal, B. D., and Broutman, L. 1.(1990) ,Analysis and Performance of Fiber
Composites, 2nd edition, Wiley.

3 Karihaloo, Bhushan L. (1995)" Fracture mechanics and structural concrete"

Harlow, Essex, England : Longman Scientific & Technical.

4 Horii, H; Hasegawa, A; and Nishino, F., (1987) "Process Zone Model and
Influencing Factors in Fracture of Concrete," in G-28, 1987, pp. 205-219.

5 Nemat-Nasser, S., and Hori, M. (1993) Micromechanics: overall properties of

Heterogeneous Materials.

6 Aveston, J., G. A. Cooper, and A. Kelly. "The Properties of Fiber Composites."

Conference Proceedings, National Physical Laboratory (!PC Science and
Technology Press Ltd). Paper 1 (1971) p. 15.

7 Mobasher, B., and Shah, S. P., "Interaction Between Fibers and the Cement
Matrix in Glass Fiber Reinforced Concrete", American Concrete Institute, ACI
SP-124, pp. 137-156, 1990

8 Jones, R.M. (1975) Mechanics of Composites Materials, McGraw Hill Book Co.

9 Mobasher, B. "Micromechanical Modeling of Filament Wound Cement-Based

Composites," ASCE, Journal of Engineering Mechanics, Volume 129, No. 4,
pp. 373-382, 2003.

10 Mobasher, B., Pivacek A., and Haupt, G. J. " Cement Based Cross-Ply
Laminates," Journal of Advanced Cement Based Materials, 1997, 6, pp. 144-
152.
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

l'2l
vt
Figure I Definition of lamina and coordinates used in generating stiffness coefficients.

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 189

Stress Damage

Figure 2. The stiffness degradation as a function of damage parameter w.

250

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
Glass Fiber Composites
200
E
zE 150
""'.3"'
-.;
c: 100
.E
0
z
50

0.001 0.002 0.003 0.004 0.002 0.003 0.004

Axial Strain, mnv·mm (A) Axial Strain, mm/mm (B)

Figure 3 Comparison of model predictions with experiments for [0/90/0] and

[0/45/-45/90/90], glass-cement systems.

o,~'"' 5 MP>t
Unidirectional 0',;",<6 MPa
50
Vr"' 9%

....... 0 <kgrc<: experiment

:~ . 0 degn.>e simul31ion
......_.. Oi90·'91Hl cxpt.-riment
.~.-f.~':. Q/90:'90:0 simulntion

(l.OOO 0.005 0.010 O.tl15 O.MC

Strain. mmimm

Figure 4 Comparison of model predictions with experiments for unidirectional, and

[0/90], glass-cement systems

Copyright American Concrete Institute

190 Mobasher
Vr=6%
w1J.- SO Softening Coeffi~;ic:m
Em= 30000 MPa
Er.,. 7000 MPa

1 0

z 4r~~~~~~--
Polypropylene
Model Simulation Fiber Composites

0.000 0.005 0.010 O.QJ5

Strain, mm/mm

Figure 5 Parametric study of effect of lamina orientation on the mechanical response.

Unidirectional, 0/90/0, and 90/0/90 glass-cement systems are compared.

[>.p~tif\t~tn,;.,
h~<:-a.::ck, Ui!t.'J'I. ttnJ :.1Qb.as.l1.:=r. l~~

\"1v;f\.",(.
f~ .·. :";ft~m MP:~.
4 P<>lypropylene
1!.~ .... N:l(l{) ~JP:,
l'w:""(I,J~ '-'t.-.(I~S
Fi!>cr Composit..-s
fl:,, ... ~·Mt.a

O.OIS
Strain, nun.imm

Figure 6. Comparison of model predictions with experiments for unidirectional

polypropylene fiber composite systems.

£
:2
l l
gli 40
<n
e
"
£"'
]
...... o dq:ree~crimc:ul
"'
;>
·:; ,·~: n dt.ogn.-esi.mul3.liou

~ ......_ Cl•9W9Ml rx(l<.-rimtnl

(~~ 0/9()/9<1:.0 simulal">n

( ..
0 4 8
Dellecti<m, mm

Figure 7 Comparison of model predictions with experiments for unidirectional,

and [0/90/90/0] glass-cement systems
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 191

E~"" lQOl\t
ft*1iHiilA
Yll
1
1+(1.{8

O.U(iOl

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
CtU\'il:ture~ Pmm

Figure 8 Comparison of model predictions for moment curvature response for

unidirectional composites containing a different volume fraction of fibers.

Copyright American Concrete Institute

192 Mobasher

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

SP-224-14

Effect of Crack Growth on the Overall

Mechanical Properties of Cement Composites

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
by M. Boulfiza and N. Banthia

Synopsis: Cement-based composites, reinforced with randomly distributed short fibers

exhibit a nonlinear behavior, called damage, which could be described in terms of
microcrack initiation, growth and coalescence leading to the creation ofmacrocracks. A
micromechanics-based continuum damage mechanics, MBCDM, model is proposed for
the prediction of the effect of initial microcrack configuration and propagation on the
macroscopic Young's modulus and thermodynamic force associated with the chosen
damage variable. Parametric studies for a number of periodic crack distributions in a two-
dimensional case have been carried out. Both unreinforced (brittle) and pitch-based
carbon fiber reinforced thin sheet cementitious materials have been considered. It is
shown that despite the relative simplicity of the damage measure used, the model was able
to capture the main effects of cracking patterns on the overall behavior of the composite.
Simulation results also reveal that, whereas the evolution of the nom1alized stiffness is
practically the same for all configurations over the entire range of damage variation, the
damage thermodynamic force is different for each case. The results predicted by the
proposed approach, appear to be consistent with experimental observations regarding the
tensile behavior ofCFRC composites.

Keywords: damage; fibers; homogenization; overall behavior; reinforced

concrete; thin sheet CFRC

193
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=University of Texas Revised Sub Account/5620001114, User=erur, ert
No reproduction or networking permitted without license from IHS Not for Resale, 01/26/2015 02:27:05 MST
Daneshlink.com
daneshlink.com

194 Boulfiza and Banthia

Mohamed Boulfiza is an Assistant Professor at the Civil Engineering Department of the
University of Saskatchewan in Saskatoon, Canada. His research interests include
integrating durability and structural design, constitutive modeling of cement-based
materials, and fiber reinforced concrete.

Nemkumar Banthia is a Professor of Civil Engineering, and Distinguished University

Scholar at the University of British Columbia in Vancouver, Canada. He serves on
various ACI, and RILEM committees and chairs the Materials Division of the Canadian
Society for Civil Engineering. In 1997, he received the American Concrete Institute's
Wason Medal for Materials Research.

INTRODUCTION

Unreinforced cementitious materials are characterized by low tensile strengths, and low
tensile strain capacities, and hence they require reinforcement. In thin sheet products
conventional reinforcing bars cannot be used, and hence, fibers constitute the primary,
and often the only, reinforcement. Typically, these materials are characterized by
relatively high fiber concentrations, exceeding 5% by volume. Here, the fibers act to
increase both the strength and the toughness of the composite. One fiber that is inert in
the cementitious environment, is not associated with any health hazards, and has been
shown to possess great potential in the production of thin products, is the carbon fiber
[2,3,4]. One of the major uses of CFRC is in thin pre-cast products such as roofing sheets,
panels, tiles, curtain walls, ferrocements, wave absorbers, permanent forms, free-access
floor panels, and 1- and L-shaped beams [3,4].

The idea of reinforcing relatively brittle building materials with fibers has been known
and practiced since ancient times, modeling the mechanical behavior of such materials is
however, not a trivial task. This is because the nonlinear behavior of these materials
depends on the type, size, distribution and orientation of microdefects, fibers and other
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

inclusions within the material. Micromechanics-based continuum damage mechanics,

MBCDM, offers a way of studying the mechanical behavior of heterogeneous materials
by considering the global macroscopic behavior as a consequence of the local one at the
scale of the heterogeneities or the different components making up the material. A
computational approach is presented for building micro-macro models for nonlinear and
randomly inhomogeneous materials such as fiber reinforced concrete, which is simpler
than the classical approach based on ergodic theory [21]. This approach which offers an
alternative to the consideration of correlation functions of increasing order is shown to be
effective for taking into account the microgeometry and microbehavior of the material
constituents when coupled with homogenization theory.

In an attempt to estimate some of the effects of the simplifying assumptions on the

determination of effective moduli and damage evolution in classical closed form
homogenization techniques [14], a numerical analysis is performed. The effect of initial
cracking patterns on the representative volume element (RYE) effective stiffness and
thermodynamic force associated with the chosen damage variable has been investigated
through parametric studies for a number of crack distributions in a two-dimensional case.

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 195

Both brittle (unreinforced) and fiber reinforced thin cementitious materials have been
considered.

RESEARCH SIGNIFICANCE

Under various external loads and environmental conditions, microcracks develop in thin
sheet cement-based composites and will eventually control the behavior of the material at
advanced stages of loading. A novel approach is proposed here to link the evolution of
the overall (macroscopic) mechanical properties of the composite considered as a
homogeneous medium with the configuration and evolution of crack type microdefects
while taking into account the properties of the fibers and the matrix. The results of this
study are expected to contribute to the development and implementation of rational
procedures for performance based design of FRC composites through a better
quantification of the effect of the material's degradation on its service performance in the
structure.

CONSTITUTIVE MODELING

The macroscopic behavior of a damaged material may be completely known once the
specific Helmholtz free energy and the damage evolution laws iJ are known. Indeed,
given the specific free energy

(1)

one can compute the associated thermodynamic forces and the stiffness tensor as follows

(jii = pOlf/
--
OBij

Olf/
s=---
8T
(2)
y = Olf/
iJ poD ..
I}

021f/
ci,.kl = P _ _..:.__
. OBijOBkl

where Bii is the strain tensor, T is temperature and DiJ is the damage tensor, CY!i is the

stress tensor, s is the entropy, YiJ is the thermodynamic force associated with damage,
and Ciikl is the 4th order stiffness tensor. The rate of change of the internal state of the
solid is governed by the damage evolution equations as independent equations of
evolution for every internal damage variable

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

196 Boulfiza and Banthia

(3)

Another alternative for writing the kinetic equations describing damage evolution may be
as derivatives of a suitably chosen potential function, making use of the generalized
normality postulate (17]. The Clausius-Duhem inequality which ensures that the first and
second principle of thermodynamics are satisfied leads to

(4)

Phenomenological continuum damage models do not reflect explicitly any of the

dissipative or energy transfers taking place at the mesoscale since they do not account for
the specific arrangement and organization of damage entities. However, they are still able
to predict the macroscopic stiffness in a rather accurate way [ 10]. In a micromechanics-
based formulation, on the other hand, damage variables are chosen in such a way that the
most important aspects of damage morphology are explicitly incorporated into the model.
Among the many damage descriptors that have been proposed in the literature one can
mention the second rank tensor proposed by Kachanov for the characterization of
deterioration due to cracking [ 10)

_ f ~ 2 (a) (a)
Di i - - L..a<a) n; n1 (5)
A a=l

where A is the area of the two dimensional RVE, a index denoting the number of the
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

crack under consideration ( 0 =5: a =5: M ), M being the total number of cracks, n(a) the
unit vector normal to crack number a, a is the half length of crack a . Similar damage
variables have been developed to account for different mechanisms taking place in a
whole variety of single phase materials and composites [1,18,19,20]. A major short-
coming in this approach, resides in the fact that macroscopic damage variables obtained
through a spatial average of individual damage entities, assumes that damage is
uniformly distributed throughout the RVE. This represents a serious limitation in
formulating damage evolution laws that include sub-RVE length scale interaction effects.

When studying the mechanical response of Fiber Reinforced Cement-based (FRC)

composites, it is important to distinguish between the case of low fiber volume fraction
composites, typically less than 2%, and high fiber volume fracti«;ms (>2%). In the former
case, the composite's response is characterized by a linearly elastic behavior followed by
a strain softening response usually associated with the creation and propagation of a
single crack around the weakest or most stressed region of the specimen. The behavior of
high fiber volume fraction composites, on the other hand, is rather fundamentally
different and characterized by the creation of multiple microcracks throughout the
specimen due to the fact that it is easier to create a new crack than to open or propagate
an existing one. This phenomenon, termed multiple cracking, leads to a strain hardening
behavior of the material [2,3,4,5,6].

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 197

A micromechanics-based model has been developed to account for the effect of fibers on
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

stress transfer across the crack in fiber reinforced cementitious composites (FRC) [6].
Unlike most models in the literature, this micromechanical model assumes that cracking
plays a major role in the composite's behavior, starting at the point of departure from
linearity up to ultimate failure and not only in the post-peak zone. The following three
stages have been considered in estimating the uncracked composite Young modulus and
the equivalent cohesive pressure acting at the lips of the crack to simulate the bridging
effect of fibers [7,8].

1- Linear Elastic Zone: uncracked case. The rule of mixtures is assumed to hold true for
the linear behavior of the composite. This theory has been shown to yield accurate results
for quasi-brittle materials at this stage of loading as long as the fiber volume fraction
remains less than I 0% [ 13].
(6)
where E, £ 117 and £ 1 are Young's moduli for the composite, the matrix and the fiber,
respectively. vm and v 1 are the matrix and fiber volume fractions.

2- Strain Hardening Zone: creation of damage by microcracking. The bridging effect of

fibers is characterized by cohesive pressure, a , that depends on the crack opening,
according to the following expression

a=
2 vJ E Ef T" max 0
(7)
(1- v1 ) rEm
where f 01 a, is the maximum shear stress at the fiber matrix interface, r is the radius of
the fiber and c5 is the crack opening.

3- Strain Softening Zone: localization of damage in a certain region due to coalescence of

micro-cracks leading to the creation of a major crack, the propagation of which will
eventually lead to ultimate failure. The fiber bridging effect at this advanced stage of
loading is represented by the following cohesive pressure
16 v1
a= r(o) ;rrd1 [L~-8L1 8+12 8 2 ] (8)
3 AI L 1
where A 1 is the cross sectional area of the fiber, L 1 is the fiber length, r(c5) is the
frictional shear bond strength at the fiber-matrix interface, and d 1 is the fiber's diameter.
It is worth noticing that the models presented here have been developed for straight fibers
and strictly speaking should not be applied when dealing with deformed fibers. However
the models could be applied to most straight fibers regardless of the nature of the material
as long as they are straight and do not exhibit excessive deformation when loaded [6].

Copyright American Concrete Institute

198 Boulfiza and Banthia

NUMERICAL MICRO-MACRO APPROACH FOR FRC COMPOSITES

The nonlinear response of heterogeneous engineering materials depends on the type

(crack, void, inclusion, etc.), size, distribution and orientation of microdefects in the
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

materials. With an increasing load, microdefects in a material evolve, and the effects of
interaction among the microdefects become dominant, leading to the creation of
macrodefects that will ultimately govern the macroscopic failure. Previous
micromechanics-based theoretical investigations, have revealed that the dominant
mechanism of localization is caused by interaction among microdefects [14,15].
Practically, the damage growth kinetic equations are usually postulated based on
empirical evidence or computational suitability. Typically, a number of simplifYing
assumptions are often made to make the problem tractable. Usually, these simplifYing
assumptions have the inconvenience of restricting the validity of the analysis to the dilute
damage case under a single specific type of loading.

The micromechanics-based models developed in the previous section are used here as
cohesive crack pressures to study the effect of fiber content and cracking configuration
on the thermodynamic force associated with the macroscopic damage variable and the
overall Young's modulus for CFRC materials. The micromechanics-based cohesive
pressure model can account for the effect of fibers on the fracture behavior of FRC
materials. The equivalence between the real cohesive crack and the model is shown in
figure I. The fiber bridging effect is modeled as a cohesive pressure that acts on the lips
of the crack and depends on its opening.

In the finite element implementation, interface elements are introduced along the crack
lips to simulate the cohesive crack. The effect of the fiber type and volume fraction is
accounted for through its specific a(J) constitutive law [6]. Equations 7 and 8 were
used to represent the cohesive pressure acting on the lips of the crack whereas equation 6
was used for the uncracked area of the RVE.

The results of a numerical simulation using the finite element model for the pure tensile
behavior of a typical RVE of a fiber reinforced specimen in the strain-hardening regime
are shown in Figure 2. Indeed, at this stage of loading the phenomenon of multiple
cracking occurs, for the fiber volume fractions considered in this investigation, and many
micro-cracks develop at different locations of the RVE prior to their evolution and
coalescence of some of them to form macrocracks. As can be seen from the stress
distribution in Figure 2, very little interaction exists between the different microcracks
prior to the formation of enough cracks in a given area of the RVE the coalescence of
which leads to crack localization.

Figure 3 illustrates the difference between a dilute case of cracking as represented by

microcracks I , 2, and 3 on the one hand, and the case of interacting cracks in the central
area of the RVE. The numerical simulation clearly shows that, whereas cracks 1, 2, 3
have no influence on each other nor on the cracks in the central area, the central cracks
have a strong influence on each other and could ultimately lead to the creation of a

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 199

macrocrack (represented by a white dashed line in the central part of the specimen), by
microcrack coalescence, the propagation of which could lead to ultimate failure.

In an attempt to estimate some of the effects of the simplifying assumptions on the

determination of effective moduli and damage evolution in classical closed form
homogenization techniques, a numerical analysis is performed in this section. The effect
of iniiial cracking patterns on the RYE-effective stiffness and thermodynamic force
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

associated with the chosen damage variable has been investigated through parametric
studies for a number of periodic crack distributions of parallel cracks in a two-
dimensional case. Both brittle and fiber reinforced cementitious materials have been
considered. Overall moduli and thermodynamic forces associated with damage were
calculated at each damage increment within a numerical simulation of evolving cracks.

The aim of this simulation is to elucidate:

- The effect of the relative size of cracks length with respect to the RYE size on the
effective stiffness and thermodynamic forces;
- The inability of low order damage measures to capture the difference between crack
distributions that yield markedly different global behaviors.

Periodic boundary conditions were applied to simulate a repeating mesostructure [14]. It

is interesting to note that periodicity of the RYE does not necessarily mean periodicity of
inclusions' distribution within the RYE. One can easily imagine the genesis of a
macroscopic material by joining together a large number of RYE's containing randomly
distributed defects or inclusions. A displacement u 2 was applied in the x 2 direction to
the upper RYE boundary. Cracks whose propagation criteria are satisfied are extended in
a self-similar manner. The propagation criterion is K 1 = K 1c for traction-free cracks
and 0'1 = /, for the cohesive cracks, with 0'1 being the tensile principal stress and /,
the tensile strength of the material.

The effective Young's modulus and damage thermodynamic force together with local
driving forces, were computed at each stable damage configuration. Damage distribution
is characterized by the crack density tensor in equation (5) which gives a typical low
order damage representation [ 11 ]. For the particular configuration considered in this
study, this tensorial damage variable has only one non-zero component given by
} N
D =D,-- =-"a.
A L.... J
(9)
J~l

and the damage thennodynamic force, also has only one non zero component,

y = y22 = - p - -
alf/ (10)
8D 12
An expression for Young's modulus E 2 in the x 2 direction can be derived using
Hooke's law for orthotropic media and associated symmetry properties [ 12] to yield

Copyright American Concrete Institute

200 Boulfiza and Banthia

&22 Ej - a22 E2 + v; Eo a22 = 0 (I 1)

where E 0 and v0 are the initial Young's modulus and Poisson's ratio before cracking.
&22 and (f22 are the average normal strain and stress, acting on the RVE boundary in

the x 2 direction.
Continuous damage evolution in a damaged RVE may be numerically simulated through
a sequence of N steps of damage increments associated with N+ 1 stable damage states,
D(i) , where i=O, .. .. .. ,N. Each stable damage state, DUl, has a strain energy density
threshold, w<il, necessary for damage evolution to take place under a given load. If the
strain density is chosen to be the thermodynamic potential, w(i) = lf/(i), then equation
( 10) can be approximated using the three-point formula [9]

. "'(i+l) - "'(i-1)
y(z) =- Y' Y'
(12)
P D(i+l) - n<H>

where i denotes the i'" stable damage state, Y the damage thermodynamic force.
Equations ( 11) and ( 12) may therefore be used to compute the averaged modulus and the
damage thermodynamic force.

RESULTS AND DISCUSSION

Uniformly distributed crack patterns are used to illustrate the effect of the relative
distribution of crack lengths on the macroscopic Young's modulus and damage
thermodynamic force for a given crack density. Figure 4 shows three uniform crack
distributions over the RVE consisting of one, four and sixteen cracks, respectively. Each
distribution has the same initial normalized horizontal and vertical spacing between
neighboring cracks, whereas the initial crack length of the second and third are about 1h
and~ of that ofthe first configuration.

Simulation results clearly show on Figure 5 that for a given damage, as represented by
the crack density, the macroscopic Young's modulus (obtained using equation 11) is
independent of the considered cracking configurations. On the other hand, Figure 6
reveals that unlike the evolution of the normalized stiffuess E 2 / E 0 , the damage
thermodynamic force (obtained using equation 12) is different for each case. Indeed, as
can be easily seen on Figure 6, the normalized thermodynamic force, Y/YREF , necessary
for damage evolution decreases as the characteristic crack size of the distribution
increases, where Y is the damage thermodynamic force and YREF is the thermodynamic
force associated with the initial damage (crack density) shown in pattern III chosen as a
reference. Despite the difference in the value of the normalized damage thermodynamic
force, the three configurations show, however, a similar tendency in that, Y/YREF
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 201

decreases rapidly with in-creasing damage, before assuming an asymptotic value which is
crack size dependent. This might be caused by the increasing interaction between parallel
cracks as they evolve within the RVE.
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

A physical interpretation of the aforementioned results could be achieved if the damage

thermodynamic force is thought of as the critical energy threshold per increment of
damage that must be exceeded for deterioration to occur. The thermodynamic force
associated with the continuum scalar damage variable has been shown to play the
equivalent role of the critical strain energy release rate in fracture mechanics [6]. Having
this analogy in mind, one could say that configurations with larger cracks reach a critical
value and evolve at lower load levels and lower strain energy release rates than
configurations with smaller cracks for the same damage state (crack density), despite the
fact that they have the same macroscopic stiffness.

Given the independence of the macroscopic Young's modulus on the cracking

configuration, together with the length computations required to simulate the case of
multiple cracking in reinforced concrete (using cohesive cracks), only pattern I on Figure
4 has been considered for investigating the effect of fiber volume fraction on the
macroscopic stiffness. Cohesive crack analyses have been carried out for three volume
fractions of pitch-based carbon fiber reinforced composites.

Figure 7 shows that for a given damage extent, the normalized macroscopic stiffness,
£ 2 /£0 , increases with fiber content. Also, for a given stiffness, the damage that can be
sustained by the material increases. It is worth noticing that beyond a fiber volume
fraction of 5%, increasing fiber content does not lead to any significant enhancements of
the materials behavior. This is consistent with experimental observations for the tensile
beha\'ior of the CFRC composites shown in Figure 8 where one can easily see that the
improvement achieved when going from v1 = 4% to v1 = 5% is not as significant as
the improvement obtained when going from v1 = 3% to v = 4%. Although the fiber
1
length used in the simulation (3mm) was different from the fiber length used in the
experiment shown (I 0 mm), it nonetheless appears that regardless of the fiber length,
there is no significant improvement beyond a certain fiber volume fraction.

In practice, the theory presented in this paper can be very useful in assessing the effects
of different fiber types and bond properties on the overall performance of different FRC
formulations as measured by the evolution of the macroscopic Young's modulus and the
continuum damage variable representing the evolution of microcracks in the FRC
composite. Another interesting application where the proposed procedure may be used is
in the area of predicting the nonlinear response of FRC structures by providing
macroscopic (or averaged) material properties needed by a traditional numerical model
for structures such as a finite element model. Indeed, using the proposed approach, one
would not need to worry about the details of composite nature of the FRC material, which
have been hidden form the structural engineer through the upscaling nature of the
procedure. This makes the analysis simpler while allowing an indirect link to the
composite constitution of the material.

Copyright American Concrete Institute

202 Boulfiza and Banthia

CONCLUSIONS

The following conclusions can be drawn:

- A micro-macro approach has been presented for the evaluation of the macroscopic
stiffness and damage in cement-based composites reinforced with randomly distributed
short fibers.
- The macroscopic stiffness has been shown to be insensitive to the considered cracking
configurations, whereas the thermodynamic force associated with damage has been found
to be sensitive to it.
- Despite the relative simplicity of the damage measure used, the model was able to
capture the main effects of crack patterns on the overall behavior of the composite.
- Improvements in the material's stiffness as a function of damage, appear to be rather
small beyond a fiber volume fraction of 5%;

ACKNOWLEDGEMENTS

The authors would like to express their gratitude to The National Research Council of
Canada for their financial support to the current project.

REFERENCES

I. Allen, D.H., Harris, C.E., and Groves, S.E. 1987. A Thermomechanical Constitutive
Theory for Elastic Composites with Distributed Damage. 1: Theoretical
Development. Int. J. Solids and Struct., Vol. 23, , pp. 1310-1318

2. Bentur, A. and Mindess, S. 1990. Fiber reinforced cementitious composites, London:

Elsevier.

3. Banthia, N. and Mindess, S., (Eds), 1995. Fiber Reinforced Concrete: Modem
Developments. Second University-Industry Workshop on Fiber Reinforced Concrete,
Toronto, Canada.

4. Banthia, N., Bentur, A. and Mufti, A., (Eds.), 1998. Fiber Reinforced Concrete:
Present and Future, Canadian Society for Civil Engineering, Montreal, Canada.

5. Boulfiza, M., Banthia, N., and Sakai, K. 2000. Application of Continuum Damage
Mechanics to CFRC Composites. ACI Materials Journal, Vol. 97, No.3, pp. 245-
253.

6. Boulfiza, M. 1998. Constitutive Modeling of Fiber Reinforced Cement Composites.

Ph.D. Thesis, University of British Columbia, Canada.

7. Boulfiza, M., Banthia, N. 2002. A discrete crack analysis for high fiber volume
fraction cement-based composites: part 1, constitutive modeling. Damage & Fracture
Mechanics 2002, Maui, Hawaii, 15-17 October.

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 203

8. Boulfiza, M., Banthia, N. 2002. A discrete crack analysis for high fiber volume
fraction cement-based composites: part 2, numerical implementation. Damage &
Fracture Mechanics 2002, Maui, Hawaii, 15-17 October.

9. Hoffman, J.D. 1992. Numerical Methods for Engineers and Scientists. New York,
McGraw-Hill.

I 0. Kachanov, L. M. I980. A Continuum Model of Medium with Cracks. Journal of

Engineering Mechanics, Vol. I 06, pp. I 039-1051.

11. Kachanov, L. M. 1992. Effective Elastic Properties of Cracked Solids: Critical

Review of Some Basic Concepts. Appl. Mech. Rev., Vol. 45, pp. 304-335.

12. Lemaitre, J., Chaboche, J.L. 1990. Mechanics of Solid Materials. Cambridge
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

UniYersity Press.

13. V.C. Li, C.K. Y. Leung, 1992. Steady-State and Multiple Cracking of Short Random
Fiber Composites. J. Engng Mech., Vol. 118, No. I I, pp. 2246-2264.

14. Nemat-Nasser, S., and Hori, M. 1999. Micromechanics: Overall Properties of

Heterogeneous Materials. North-Holland; 2nd edition.

15. Nim1alendran, S., Horii, H. 1992. Analytical modeling ofmicrocracking and

bridging in the fracture of quasi-brittle materials. J. Mech. Phys. Solids Vo140, N 4.

16. Okui, Y., 1-lorii, H., and Akiyama, N. 1993. A Continuum Theory for Solids
Containing Microdefects. Int. J. Engng Sci. Vol. 31, N° 5, pp. 735-749.

17. Rice, .I.R. 1971. Inelastic Constitutive Relations for Solids: An Internal-Variable
Theory and its Application to Metal Plasticity. J. Mech. Phys. Sol-ids, Vol. 19, pp.
433-455.

18. Swenson, D.V. 1985. Modeling Mixed-Mode Dynamic Crack Propagation Using
Finite Elements. Ph.D. Thesis, Corne]) University.

19. Talreja, R. 1990. Internal Variable Damage Mechanics of Composite Materials. In

Yielding Damage and Failure of Anisotropic Solids, ed. J.P. Bohler, London:
Mechanical Engineering Publications, pp. 509-533.

20. Talreja, R. I 991. Continuum Modeling of Damage in Ceramic Matrix Composites.

Mech. Mater., Vol. 12, pp. 165-180.

21. Torquato, S. 2002. Random heterogeneous materials: microstructure and

macrostructrue properties. Springer.

Copyright American Concrete Institute

204 Boulfiza and Banthia

Table 1. Material Parameters for the Considered Mortar and Fibers.
Matrix Properties (Mortar)
Material Property Units Value
Young·s modulus MPa 9385
Poisson ratio 0.2
WIC 0.35
SIC 0.5
SF/C 0.2
Fiber Properties
Material Property Units Value
Fiber Type Pitch-based
Diameter (pm) 18
Specific Gravity 1.7
Young·s modulus GPa 30
Tensile strength MPa 590

Figure 1- Schematic representation of cracking in FRC composites a) real crack,

b) physical model

~-
XLx~t=t=t=f=t=t=t
I

I I
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

·+
. . :,~,i=t=~=+=~~"~+-······-
i..__.....-.::=_!
I._. . . . . .
Figure 2- A typical RVE in the strain hardening regime.

Copyright American Concrete Institute

Thin Reinforced c,p~ent-Based Products 205

Interacting cnackl

c:::::::~
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Figure 3- Main difference between dilute and in!eracting cracks in an RYE.

Patlernl Pll~Wrnll PQIU!m/11

I hlw I I I

I Al"'RI'E 0.3 0.14 0.14

Pa«eml Puttern II Patleml/1

Figure 4-Crack patterns considered to investigate the effect of cracking on the overall
properties of the :RYE.

1.0 '"'"''""""''""'''""""""'·-""""""'"····-··--·--..,.·-·:-·--··-··--·-·--,
0.9
~ 0.8 ~Pattern I

;a 0.7 ..... Pattern II

• Pattern ,II.
I§= 0.6
~ 0.5
a: 0.4
l
.!:!
0.3
~ 0.2
0.1
0.0
000 0.05 0.10 0.15 0.20 0.25
Damage (Crack Density)

Figure 5-Effect of cracking configuration and exten~ on the tnacroscopic Young's

modulus for unreinforced matrices,

Copyright American Concrete Institute

206 Boulfiza and Banthia

1.6 · r - - - - - r - - - - - - - - r - - - - - , - - - - - - ;

1.4 . :+ullifi:im!i '

\4-Unlform II '
1.2 '-+-Unl,fCJim Ill
1.0 .. .. ':~ .

f ~ L~-···~·
~· : = · ~·= ·- ~·; : ·~ ·- ·:j~ -~· ·:~.-;=-:!·~: -: · =·: !· ~_J
0 5 10 15
.... . . .
25

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---
Damage(%}

Figure 6- Effect of cracking configuration and extent on the thermodynamic force

associateq with the macroscopic damage variable.

1 r············· ................................-., ..................................................,.....................,...,........................,

I
o.9 i
0.8,~-- ·'+'Vi~%·:y . !
' : i
: 0.7 ;..... vt=3%~.
, ;...-vt:s%
i, 0.6 ..... i~\f,f:=!'*.>

i 0.5
1 0.4
j o.3 l
0.21
o.1 II
0''-.- - - -
0 0.1 0.2 0.3 0.4 0.5 0.6
Damage (cracJt. density)

Figure 7-Effect of fiber content and extent of damage on the macroscopic Young's
modulus.
7 .......... -·~··..---,~·-----.-··-·--·-~-·---;"-----. _ ,__ l
...... ···~ .. -+-Vf=O%
6

14
....
,!3
21
1 I
0 0.1 0.2 0.3 OA 0.5
Strain(%)
Figure 8- Tensile stress strain behavior of a CFRC composite as a function of fiber
volume fraction.

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 207

CONVERSION FACTORS-INCH-POUND TO SI (METRIC)*
To convert from to multiply by

Length
inch millimeter (mm) 25.4Et
foot metcr(m) 0.3048E
yard meter (m) 0.9144E
mile (statute) kilometer (km) 1.609

Area
square inch square centimeter (cm 2) 6.451
square foot 2 0.0929
square meter (m )
square yard 2 0.8361
square meter (m )

Volume (capacity)
ounce cubic centimeter (cm 3 ) 29.57
gallon cubic meter (rn\j: 0.003785
cubic inch cul>ic centimeter (em·') 16.4
cubic foot cubic meter (m 3) 0.02832
cubic yard cul>ic meter (m\i: 0.7646

Force
kilogram-force newton(N) 9.807
kip-force newton (N) 4448
pound-force newton (N) 4.448 --``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Pressure or stress
(force per area)
kilogram-force/square meter pascal (Pa) 9.807
kip-force/square inch (ksi) megapascal (MPa) 6.895
newton/square meter (N/m 2 ) pascal (Pa) I.OOOE
pound-force/square foot pascal (Pa) 47.88
pound-force/square inch (psi) kilopaseal (kPa) 6.895

Bending moment or torque

inch-pound-force newton-meter (Nm) 0.1130
foot-pound-force newton-meter (Nm) 1.356
meter-kilogram-force newton-meter (Nm) 9.807

Copyright American Concrete Institute

208 Conversion Factors-Inch-Pound to Sl (Metric)

To convert from to multiply by
Mass
ounce-mass (avoirdupois) gram (g) 28.34
pound-mass (avoirdupois) kilogram (kg) 0.4536
ton (metric) megagram (Mg) J.OOOE
ton (short, 2000 Ibm) megagram (Mg) 0.9072

Mass per volume

pound-mass/cubic fool kilogram/cubic meter (kg/m 3) 16.02

pound-mass/cubic yard kilogram/cubic meter (kg/m 3) 0.5933

pound-mass/gallon kilogram/cubic meter (kgtm 3) 119.8

Temperature§
deg Fahrenheit (F) deg Celsius (C) lc =(tp - 32)/1.8
<.leg Celsius (C) deg Fahrenheit (F) tp = 1.8'c + 32

* This selected list gives practical conversion factors of units found in concrete technology. The reference
source for information on SI units and more exact conversion factors is "Standard for Metric Practice" ASTM E
380. Symbols of metric units are given in parentheses.
t E indicates that the ractor given is exact.
:j: One liter (cubic decimeter) equals 0.001 m 3 or 1000 cm 3.
* These equations convert one temperature reading to another and include the necessary scale corrections. To
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

convert a difference in temperature from Fahrenheit to Celsius degrees, divide by 1.8 only, i.e .• a change from 70
to 88 F represents a change of 18 For 18/1.8 = I 0 C.

Copyright American Concrete Institute

Thin Reinforced Cement-Based Products 209

Index

A E
abrasion, 71 engineering properties, 1
aesthetics, 71 expanded polystyrene, 101
analysis. 33
AR glass fibers, 1 F
AR mats, I facades, 55
AR meshes. I fatigue behavior, 89
aramid. 21 ferrocement, 127
fiber, 161
B fibers, 179, 193
Banthia. N., 161, 193 fiber cement, I
Barle, M., 45 fiber-reinforced, 71
Blain-Cosgrove, E., IOI fiber reinforced cement, 1 13
Boulfiza, M .. 193 fiber-reinforced cement board, 145
Brameshuber, W., 45 fiber-reinforced concrete, 21
Bruckermann, 0., 33 fiber reinforcement, I
finite elements, 33
c flexural and compressive strength, I13
carbon fibers, I 13 form work, 45
cement. I6I. 179 frameless housing, 101
cement composite, 89 freeze-thaw durability, 145
cementitious composites, 179 full scale wall tests, 101
centrifugation, I I 3
cladding pands, 55 G
composite, 7 I Gilbert, G. T., I
concrete, 45, I 79 glass concrete, 2I
cracking, I 79 gravity load, 101
curing, 127 Gries, T., 45
curtain wall, 55
H
D Hatschek process, I45
damage, 193 Hegger, J., 33, 45,55
decorative elements, 1 Hesselbarth, D., II3
decretized, I 27 homogenization, 193
ductile, 7I
durability, I, 71
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

210 Index
s
impact, 71 sandwich building system, 101
impermeability, 71 Schneider, H., 55
interlaminar bond strength, 145 seismic resistance, 101
Shah, S. P., 145
K Shao, Y., 101
Kaufmann, J., I 13 Sherif, A., 33, 55
Konrad, M., 33 Shum, C. H., 127
--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Koster, M., 45 Sorelli, L., 161

Kri.iger, M., 45 static behavior, 89
Kuder, K. G., 145
T
M Tam, C. T., 127
manufacturing methods, 1 Tan, S. T., 127
material model, 33 temperature, 127
Meyer, C., 21 Teo, C. P., 127
microwave, 127 textile reinforced concrete, 33, 45
Mobasher, B., 179 textile reinforcement, 21, 55
modular buildings, 1 thin cement products, 101
Molter, M., 55 thin cementitious products, 1
multi-level, 33 thin sheet CFRC, 193
multiscale fiber reinforcement, 89 thin sheets, 21

0 u
Ong, K. C. G., 127 UHPC, 71
overall behavior, 193 usage-life, 71

p v
Parant, E., 89 Vi Ikner, G., 21
Perry, V., 71 Voss, S., 45, 55
Plizzari, G. A., 161
polymeric fibers, 1 13 w
porosity, 145 wall panels, 1
pressure, 145 wind resistance, 101
prestressed concrete, 21 Wong, L. H. J., 127

R z
reinforced concrete, 193 Zakariasen, D., 71
reinforcement, 161
Reinhardt, H.- W., 45
Robinson, B., 101
Rossi, P., 89

Copyright American Concrete Institute

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

Copyright American Concrete Institute

--``,```,`,`,`,,,,`,`````,`,,`,,-`-`,,`,,`,`,,`---

American Concrete Institute•

Advancing concrete knowledge

Copyright American Concrete Institute

The Kama Sutra in 200 Positions
84% (113)
The Kama Sutra in 200 Positions
215 pages
Jouissance Club Livre
100% (5)
Jouissance Club Livre
258 pages
The Complete French Conjugation Course - Cover
100% (5)
The Complete French Conjugation Course - Cover
447 pages
DUNOD DCG 2 Droit Des Sociétés
94% (16)
DUNOD DCG 2 Droit Des Sociétés
497 pages
Vocabulaire Progressif Du Français Corrigés
100% (11)
Vocabulaire Progressif Du Français Corrigés
65 pages
Corrigés Grammaire Progressive Du Français Niveau Perfectionnement PDF
100% (7)
Corrigés Grammaire Progressive Du Français Niveau Perfectionnement PDF
84 pages
200 Common French Verbs
88% (8)
200 Common French Verbs
3 pages
Complete French (Learn French With Teach Yourself)
92% (64)
Complete French (Learn French With Teach Yourself)
509 pages
Collins Easy Learning French Grammar (Gnv64)
97% (58)
Collins Easy Learning French Grammar (Gnv64)
162 pages
Easy French Step-by-Step PDF
100% (23)
Easy French Step-by-Step PDF
399 pages
Abc Dalf C1-C2 Livre @TefTeacher
100% (8)
Abc Dalf C1-C2 Livre @TefTeacher
330 pages
Colloquial French - The Complete Course For Beginners (Colloquial Series) PDF
100% (11)
Colloquial French - The Complete Course For Beginners (Colloquial Series) PDF
287 pages
Advanced French Grammar PDF
97% (35)
Advanced French Grammar PDF
717 pages
Le Petit Prince
100% (15)
Le Petit Prince
57 pages
Conservation of Bridges-Graham Tilly
No ratings yet
Conservation of Bridges-Graham Tilly
417 pages
Short Stories FR Intermediate PDF
100% (7)
Short Stories FR Intermediate PDF
185 pages
365 Days of French Expressions Without Audio
100% (3)
365 Days of French Expressions Without Audio
237 pages
French Grammar Basic
60% (10)
French Grammar Basic
16 pages
French For Beginners - The Best Handbook For Learning To Speak French PDF
100% (10)
French For Beginners - The Best Handbook For Learning To Speak French PDF
101 pages
French Made Simple
93% (14)
French Made Simple
196 pages
French Workbook
100% (10)
French Workbook
288 pages
French Vocabulary
60% (5)
French Vocabulary
31 pages
French Vocabulary
88% (8)
French Vocabulary
5 pages
Le Français en Images
91% (11)
Le Français en Images
86 pages
Infinitive: Simple Past
100% (14)
Infinitive: Simple Past
4 pages
A Definitive Guide To French Verbs
83% (6)
A Definitive Guide To French Verbs
106 pages
233 Complete French Grammer - Sample
82% (11)
233 Complete French Grammer - Sample
20 pages
Beginner French
100% (17)
Beginner French
252 pages
Aci sp-210-2003
No ratings yet
Aci sp-210-2003
255 pages
Aci sp-211-2003
No ratings yet
Aci sp-211-2003
370 pages
Aci sp-228-2005
No ratings yet
Aci sp-228-2005
1,562 pages
Aci sp-218-2004
No ratings yet
Aci sp-218-2004
211 pages
Aci sp-238-2006
100% (1)
Aci sp-238-2006
499 pages
2013 Young Researchers Conference Proceedings PDF
100% (1)
2013 Young Researchers Conference Proceedings PDF
142 pages
GEO Shangh Art.4.korea A PDF
No ratings yet
GEO Shangh Art.4.korea A PDF
6 pages
Punching Shear Resistance of Slab
No ratings yet
Punching Shear Resistance of Slab
8 pages
Fang, Lundgren, Chen, Zhu-Corrosion Influence On Bond in Reinforced Concrete
No ratings yet
Fang, Lundgren, Chen, Zhu-Corrosion Influence On Bond in Reinforced Concrete
9 pages
Info de Macro Fibra Sintetica de Concrete Org Uk
No ratings yet
Info de Macro Fibra Sintetica de Concrete Org Uk
1 page
Rational Model For Calculating Deflection of Reinforced Concrete Beams and Slabs
No ratings yet
Rational Model For Calculating Deflection of Reinforced Concrete Beams and Slabs
11 pages
1963 - Bresler Scordelis - Shear Strength of RC Beams PDF
No ratings yet
1963 - Bresler Scordelis - Shear Strength of RC Beams PDF
24 pages
Buettner, Becker - PCI Manual For The Design of Hollow Core Slabs (2000)
No ratings yet
Buettner, Becker - PCI Manual For The Design of Hollow Core Slabs (2000)
141 pages
BS 2830
No ratings yet
BS 2830
46 pages
Oresund 2
No ratings yet
Oresund 2
29 pages
Proceeding of 13th Nordic Steel 2015 Construction Conference PDF
No ratings yet
Proceeding of 13th Nordic Steel 2015 Construction Conference PDF
254 pages
Road and Airport Engineering: New Frontiers
No ratings yet
Road and Airport Engineering: New Frontiers
362 pages
Dissertation RJM Pijpers September 2011
No ratings yet
Dissertation RJM Pijpers September 2011
294 pages
24 - Stress Concentration PDF
100% (1)
24 - Stress Concentration PDF
9 pages
99 MN03
No ratings yet
99 MN03
276 pages
2011 Bridge Sarkady
No ratings yet
2011 Bridge Sarkady
17 pages
Concrete Against Explosion2
No ratings yet
Concrete Against Explosion2
145 pages
SRIA TN42-2 Welding of Reinforcement
No ratings yet
SRIA TN42-2 Welding of Reinforcement
2 pages
Using Fracture Grouting To Lift Structures in Clayey Sand
No ratings yet
Using Fracture Grouting To Lift Structures in Clayey Sand
8 pages
Design of ACI-type Punching Shear Reinforcement To Eurocode 2
No ratings yet
Design of ACI-type Punching Shear Reinforcement To Eurocode 2
15 pages
Main Study Final
No ratings yet
Main Study Final
184 pages
Prefabrikaty
No ratings yet
Prefabrikaty
25 pages
13.shear Load Transfer For Rock-Socketed Drilled Shafts Based On Borehole Roughness and Geological Strength Index (GSI) PDF
No ratings yet
13.shear Load Transfer For Rock-Socketed Drilled Shafts Based On Borehole Roughness and Geological Strength Index (GSI) PDF
14 pages
State-of-the-Art of Integral Abutment Bridges: Design and Practice
No ratings yet
State-of-the-Art of Integral Abutment Bridges: Design and Practice
10 pages
Bolted Steel Slip-Critical Connections With Fillers
No ratings yet
Bolted Steel Slip-Critical Connections With Fillers
10 pages
Prec 38
No ratings yet
Prec 38
6 pages
Hif 20062
No ratings yet
Hif 20062
23 pages
نسخة من 234R-06 Guide for the Use of Silica Fume in Concrete
No ratings yet
نسخة من 234R-06 Guide for the Use of Silica Fume in Concrete
63 pages
Heinemann Historic Concrete From Concrete Repair To Concrete Conservation
No ratings yet
Heinemann Historic Concrete From Concrete Repair To Concrete Conservation
560 pages
Chimney Journal 1
No ratings yet
Chimney Journal 1
6 pages
Analysis of Segmentally Constructed Prestressed Concrete Bridges Using Hexahedral Elements With Realistic Tendon Profiles
No ratings yet
Analysis of Segmentally Constructed Prestressed Concrete Bridges Using Hexahedral Elements With Realistic Tendon Profiles
9 pages
Design of Hybrid Concrete Buildings
No ratings yet
Design of Hybrid Concrete Buildings
74 pages
By, Akshata A Mulgund
No ratings yet
By, Akshata A Mulgund
153 pages
Analysis of Full Depth Precast Concrete Bridge Deck Panels
No ratings yet
Analysis of Full Depth Precast Concrete Bridge Deck Panels
12 pages
Seepage Behavior and Grouting Effects For Large Rock Caverns
No ratings yet
Seepage Behavior and Grouting Effects For Large Rock Caverns
8 pages
Precast Segmental Bridge Construction in Seismic Zones
No ratings yet
Precast Segmental Bridge Construction in Seismic Zones
15 pages
AP-T313-16 Realising 100-Year Bridge Design Life in An Aggressive Environment Literature Review
No ratings yet
AP-T313-16 Realising 100-Year Bridge Design Life in An Aggressive Environment Literature Review
144 pages
Abu Dhabi Proceedings 3-12-09
No ratings yet
Abu Dhabi Proceedings 3-12-09
361 pages
Perfomance Evaluation of Rebar in Accelerated Corrosion
No ratings yet
Perfomance Evaluation of Rebar in Accelerated Corrosion
6 pages
Durham E-Theses: Tension Stiffening in Reinforced Concrete Members With Large Diameter Reinforcement
No ratings yet
Durham E-Theses: Tension Stiffening in Reinforced Concrete Members With Large Diameter Reinforcement
119 pages
Hebbec PDF
No ratings yet
Hebbec PDF
15 pages
Multipliers For Deflections in Reinforced Concrete Flat Slabs
No ratings yet
Multipliers For Deflections in Reinforced Concrete Flat Slabs
10 pages
Jennifer Dietrich MSC DissertationAnalysis and Design of PDF
No ratings yet
Jennifer Dietrich MSC DissertationAnalysis and Design of PDF
238 pages
Ultra High Performance Concrete A State of The Art Report For The
No ratings yet
Ultra High Performance Concrete A State of The Art Report For The
176 pages
Morgan, H. (1961) - 'A Contribution To The Analysis of Stress in A Circular Tunnel.' Geotechnique, 11
No ratings yet
Morgan, H. (1961) - 'A Contribution To The Analysis of Stress in A Circular Tunnel.' Geotechnique, 11
10 pages
Design and Construction of New Zealand's First 1825 Super Tee Beam
100% (1)
Design and Construction of New Zealand's First 1825 Super Tee Beam
16 pages
S-N Curves For Fatigue of Bond in Reinforced Concrete Structures Under
No ratings yet
S-N Curves For Fatigue of Bond in Reinforced Concrete Structures Under
7 pages
HAR1902
No ratings yet
HAR1902
152 pages
Thermal Property of Cement Paste
No ratings yet
Thermal Property of Cement Paste
18 pages
I-35w Highway Minnesota Bridge Collapse Report
100% (1)
I-35w Highway Minnesota Bridge Collapse Report
178 pages
Young Researchers Forum II Construction
No ratings yet
Young Researchers Forum II Construction
202 pages
Probabilistic service life model of RC structures subjected to the combined effect of chloride-induced corrosion and cyclic loading
From Everand
Probabilistic service life model of RC structures subjected to the combined effect of chloride-induced corrosion and cyclic loading
Emilio Bastidas Arteaga
No ratings yet
Wind Farm Noise: Measurement, Assessment, and Control
From Everand
Wind Farm Noise: Measurement, Assessment, and Control
Colin H. Hansen
No ratings yet
Aci sp-219-2004
No ratings yet
Aci sp-219-2004
174 pages
Aci sp-209-2002
No ratings yet
Aci sp-209-2002
916 pages
Aci sp-215-2003
No ratings yet
Aci sp-215-2003
446 pages
Aci sp-222-2004
No ratings yet
Aci sp-222-2004
252 pages
Aci sp-221-2004
No ratings yet
Aci sp-221-2004
978 pages
Aci sp-239-2006
No ratings yet
Aci sp-239-2006
587 pages
Aci sp-226-2005
No ratings yet
Aci sp-226-2005
158 pages
Aci sp-234-2006
No ratings yet
Aci sp-234-2006
833 pages
Aci sp-231-2005
No ratings yet
Aci sp-231-2005
297 pages
Language Hacking French (Learn How To Speak French Book)
100% (6)
Language Hacking French (Learn How To Speak French Book)
622 pages
French Grammar
100% (33)
French Grammar
368 pages
C'est-A-Dire - Advanced French Conversation PDF
100% (9)
C'est-A-Dire - Advanced French Conversation PDF
316 pages
An Energy Approach To Quantification of Adhesion Strength From Critical Loads in Scratch Tests
No ratings yet
An Energy Approach To Quantification of Adhesion Strength From Critical Loads in Scratch Tests
7 pages
Fox Running Manual: Wed, 13 May 2020 12:46:25 +0000 Valid at Time of Download
No ratings yet
Fox Running Manual: Wed, 13 May 2020 12:46:25 +0000 Valid at Time of Download
14 pages
Mid Term Exam Chemistry Paper 2023-24
No ratings yet
Mid Term Exam Chemistry Paper 2023-24
3 pages
CLB21003 Process Heat Transfer Experiment 1: Heat Conduction (Discussion) Lab Report
100% (1)
CLB21003 Process Heat Transfer Experiment 1: Heat Conduction (Discussion) Lab Report
3 pages
Molecule Shapes Activity
No ratings yet
Molecule Shapes Activity
7 pages
NAX Premila 2K Polyester Putty-MSDS
No ratings yet
NAX Premila 2K Polyester Putty-MSDS
6 pages
Assignment 1 (Topic 1 and Topic 2) Sem 2 2013-2014 (Latest)
No ratings yet
Assignment 1 (Topic 1 and Topic 2) Sem 2 2013-2014 (Latest)
7 pages
Microstructure and Mechanical Properties of Rheo-Diecasted A390 Alloy
No ratings yet
Microstructure and Mechanical Properties of Rheo-Diecasted A390 Alloy
4 pages
g250 1
No ratings yet
g250 1
2 pages
Squid PPT 1
No ratings yet
Squid PPT 1
76 pages
03b - Buffer Ws Answers and Titration Notes
No ratings yet
03b - Buffer Ws Answers and Titration Notes
47 pages
Chem
No ratings yet
Chem
19 pages
Structural Calculation Template
No ratings yet
Structural Calculation Template
19 pages
Presentation On Shuttering
100% (1)
Presentation On Shuttering
29 pages
Power FET (DMOS, VMOS) - Hansen and Roschen
No ratings yet
Power FET (DMOS, VMOS) - Hansen and Roschen
14 pages
SDR41 - J: Size (3')
No ratings yet
SDR41 - J: Size (3')
6 pages
Genesys LSG MSDS
No ratings yet
Genesys LSG MSDS
3 pages
The foundation engineering handbook Second Edition Gunaratne download pdf
No ratings yet
The foundation engineering handbook Second Edition Gunaratne download pdf
55 pages
Butyl Rubber IIR
No ratings yet
Butyl Rubber IIR
3 pages
Loss Characteristics of Optical Fiber
93% (15)
Loss Characteristics of Optical Fiber
9 pages
Unit IV Pilling Bedworth Ratio
No ratings yet
Unit IV Pilling Bedworth Ratio
7 pages
Overvew Column Analysis
100% (3)
Overvew Column Analysis
40 pages