effect of microfibrillar cellulose on concrete equivalent ...477428/fulltext01.pdf · as an...
TRANSCRIPT
Effect of microfibrillar cellulose on
concrete equivalent mortar fresh and hardened properties
Inverkan av mikrofibrillär cellulosa på egenskaperna hos betongekvivalent bruk i dess färska och hårdnade tillstånd
Author: Jonas Nilsson and Peter Sargenius
Principal: Swedish Cement and Concrete Research Institute, CBI
Supervisor: Peter Billberg, CBI
Ali Farhang, KTH ABE
Educational Institution: KTH, ABE, Department of Civil and Architectural Engineering
Examiner: Sven Henrik Vidhall, KTH ABE
Thesis: 15 course credits in the program Construction Engineering and
Design
Serial nr: 2011;24 ABE
Approval date: 2011-11-11
2011-11-11
Abstract
A pilot project in 2010, conducted at CBI, showed the capacity for pulp, micro fibrilars from
the forest industry to act as Viscosity Modifying Agent (VMA) in concrete. This project was,
however, too limited to find answers for optimal use of this kind of material. The forest
industry company Stora Enso wants to find out if their pulp can be used in concrete in order to
somehow improve its properties. Two micro fibrilar suspensions have been tested. The tested
fibrils are in two sizes, the finer material named MFC1 has undergone more homogenization
than the course material named MFC2. The fibrils have been evaluated in regard to how the
fibrils react with mortars in both its fresh and hardened state. Tests have been conducted on
the use of concrete equivalent mortars with a maximum aggregates size of 4 mm. Two water-
cement-ratios have been used in the tests, 0.45 and 0.60. Three different fibril dosages have
been tested, 1, 2 and 3 kg/m³. The results of these trials of cellulose fibrils has been evaluated
in respect of rheology, compressive strength, flexural strength, cracking, shrinkage, water
capillary porosity, anti-wash out resistance (underwater concrete) and as a possible surface
coverage.
The results from the trials, conducted in this report, show that an increased dosage of fibrils
leads to an increased plastic viscosity. The fibrils appear to have no effect on the flexural- and
compressive strength, and no effect on the shrinkage of the test specimens. According to our
results it is not advisable to use the fibrils for the purpose of acting as an agent for anti-
washout resistance, or as a surface coverage.
The work have been performed at Swedish Cement and Concrete Research Institute, CBI, in
Stockholm in the spring of 2011.
CBI is an institution whose mission is to create, apply and disseminate knowledge in the
concrete and rock area.
Keywords: micro fibrillar cellulose; cellulose fibres; rheology; shrinkage; yield stress; plastic
viscosity; anti-washout; cracking; flexural strength; compressive strength; water capillary
porosity.
Sammanfattning
Ett tidigare pilotprojekt har under 2010 utförts på CBI, och där undersöktes möjligheten för
cellulosafibrer från skogsindustrin att fungera som Viscosity Modifying Agent (VMA) i
betong. Utrymmet i detta projekt var dock för begränsat för att finna svar för optimal
användning av denna typ av material. Nu vill skogsindustriföretaget Stora Enso ta reda på om
massa från deras träprodukter kan användas i betong, för att på något sätt förbättra dess
egenskaper. Vi har därför provat suspensioner innehållande två olika fraktioner av
cellulosafibriller. De testade fibrerna finns i två storlekar, det finare materialet heter MFC1
och har genomgått med homogenisering än det grövre materialer som heter MFC2. Dessa två
typer har tillsats i bruk och utvärderats i hur de reagerar i både brukets färska och dess
hårdnade tillstånd. Testerna har genomförts på bruk med en maximal ballaststorlek på 4 mm.
Två vct-nivåer har använts i försöken, 0,45 och 0,60. Tre olika fibrilldoser har prövats,
nämligen 1, 2 och 3 kg/m³ fibriller. Resultaten från dessa försök av cellulosafibriller har
utvärderats med avseende på reologi, tryckhållfasthet, böjhållfasthet, sprickbildning,
kapillaritet, krympning, anti-urvaskning och som möjlig ytbetäckning.
De tester som har genomförts visar att med ökad dos fibriller ökar den plastiska viskositeten.
Fibrillerna visade sig inte ha någon effekt på böj- eller tryckhållfasheten, samt ingen effekt på
krympning av provkropparna. Testerna visar att fibrillerna inte heller agerar med någon
possitiv effekt som anti-urvaskningsmedel, eller som ett täckande ytskikt.
Försöken har genomförts vid CBI Betonginstitutet i Stockholm mellan 21 mars och 8 juli år
2011.
CBI är en institution vars uppdrag är att skapa, tillämpa och sprida kunskap inom betong och
bergområdet.
Nyckelord: mikrofibrillär cellulosa; cellulosafibrer; reologi; krympning; flytgräns; plastisk
viskositet; anti-urvaskning; sprickbildning; böjhållfasthet; tryckhållfasthet; kapillaritet
Foreword
This thesis is the final course of the Bachelor of Science program Construction Engineering &
Design, at the Royal Institute of Technology, KTH, in Stockholm. It is written by Jonas
Nilsson and Peter Sargenius who are studying the Construction Engineering and Design
program focusing on construction. The thesis has been performed in 16 weeks, with 15
corresponding course credits, in the spring semester of 2011 at the Swedish Cement and
Concrete Research Institute, CBI.
The examiner has been Sven-Henric Vidhall (teacher at KTH), supervisor Tech.Dr Ali
Farhang (Head of Department Bridges and Tunnels at Ramböll and teacher at KTH) and
supervisor Tech.Dr Peter Billberg (Senior Researcher at CBI).
First we would like to thank our extraordinary supervisor Peter Billberg who has helped us
during the whole project with information and guidance. We also would like to thank Patrick
Rogers and Carsten Vogt for assistance with both laboratory work, and with the project as
whole. We would like to thank all the other employees at CBI who helped us along the way.
Thanks to Ali Farhang for his guidance and input on this thesis, Irene Wedin at Stora Enso
and Sven Henrik Vidhall, examinator at KTH.
Contents
1. Introduction ....................................................................................................................... 1
1.1. Background .................................................................................................................. 1
1.2. Aim .............................................................................................................................. 1
1.3. Research significance .................................................................................................. 1
1.4. Methodology ................................................................................................................ 2
1.5. Limitations ................................................................................................................... 2
1.6. About Swedish Cement and Concrete Research Institute, CBI ................................... 3
2. Literature survey .............................................................................................................. 5
2.1. Concrete ....................................................................................................................... 5
2.2. Fresh concrete properties ............................................................................................. 5
2.2.1. Rheology .............................................................................................................. 5
2.2.2. Workability ........................................................................................................... 9
2.3. Hardened concrete properties ...................................................................................... 9
2.4. Fibres in concrete ....................................................................................................... 10
2.5. Underwater concrete .................................................................................................. 11
3. Methods and materials ................................................................................................... 13
3.1. Equipment .................................................................................................................. 13
3.1.1. Laboratory .......................................................................................................... 13
3.1.2. Sieves ................................................................................................................. 13
3.1.3. Mixer .................................................................................................................. 14
3.1.4. Air meter ............................................................................................................ 15
3.1.5. Concrete viscometer ........................................................................................... 15
3.1.6. Flexural strength testing machine ...................................................................... 16
3.1.7. Compressive strength testing machine ............................................................... 17
3.1.8. Free shrinkage testing apparatus ........................................................................ 17
3.1.9. Capillary porosity ............................................................................................... 18
3.1.10. Moulds ............................................................................................................ 18
3.2. Materials .................................................................................................................... 21
3.2.1. Aggregates .......................................................................................................... 21
3.2.2. Cement ............................................................................................................... 22
3.2.3. Fibrils ................................................................................................................. 22
3.2.4. Superplasticizer .................................................................................................. 23
3.2.5. Anti-washout admixture ..................................................................................... 24
3.2.6. Water .................................................................................................................. 24
3.3. Methods ..................................................................................................................... 25
3.3.1. Rheology and workability .................................................................................. 25
3.3.1.1. Mini slump flow ............................................................................................. 25
3.3.1.2. Air content ...................................................................................................... 25
3.3.1.3. Rheology ......................................................................................................... 25
3.3.2. Mechanical properties ........................................................................................ 26
3.3.2.1. Flexural strength ............................................................................................. 26
3.3.2.2. Compressive strength ..................................................................................... 26
3.3.3. Shrinkage ............................................................................................................ 26
3.3.3.1. Free shrinkage ................................................................................................. 26
3.3.3.2. Ring test .......................................................................................................... 27
3.3.3.3. Surface coverage ............................................................................................. 27
3.3.4. Wash-out test ...................................................................................................... 28
3.3.5. Capillary porosity ............................................................................................... 28
4. Mixture design and methodology .................................................................................. 31
4.1. Concrete equivalent mortar ....................................................................................... 31
4.2. Mixing sequence ........................................................................................................ 32
4.3. Characterization of fresh mortar ................................................................................ 33
4.4. Moulding of the mortars ............................................................................................ 33
4.5. After demoulding ....................................................................................................... 33
5. Results and discussion .................................................................................................... 35
5.1. Fresh mortar properties .............................................................................................. 35
5.1.1. Series 1 ............................................................................................................... 35
5.1.2. Series 2 ............................................................................................................... 39
5.1.3. Series 3 ............................................................................................................... 44
5.1.4. Series 4 ............................................................................................................... 46
5.1.5. Series 5 ............................................................................................................... 48
5.1.6. Discussion .......................................................................................................... 51
5.2. Mechanical properties ................................................................................................ 55
5.2.1. Compressive and flexural strength ..................................................................... 55
5.2.1.1. Series 1 ........................................................................................................... 55
5.2.1.2. Series 2 ........................................................................................................... 57
5.2.1.3. Series 3 ........................................................................................................... 60
5.2.1.4. Series 4 ........................................................................................................... 62
5.2.1.5. Series 5 ........................................................................................................... 64
5.2.2. Discussion .......................................................................................................... 66
5.3. Shrinkage ................................................................................................................... 68
5.3.1. Free Shrinkage .................................................................................................... 68
5.3.1.1. Series 1 ........................................................................................................... 68
5.3.1.2. Series 2 ........................................................................................................... 69
5.3.1.3. Series 3 ........................................................................................................... 70
5.3.1.4. Series 4 ........................................................................................................... 71
5.3.2. Constrained shrinkage ........................................................................................ 72
5.3.3. Surface coverage ................................................................................................ 73
5.3.4. Discussion .......................................................................................................... 73
5.4. Weight loss ................................................................................................................ 75
5.4.1. Series 1 ............................................................................................................... 75
5.4.2. Series 2 ............................................................................................................... 76
5.4.3. Series 3 ............................................................................................................... 77
5.4.4. Series 4 ............................................................................................................... 78
5.4.5. Discussion .......................................................................................................... 79
5.5. Anti washout resistance ............................................................................................. 80
5.5.1. Discussion .............................................................................................................. 81
5.6. Capillary porosity ...................................................................................................... 81
6. Conclusions ...................................................................................................................... 85
7. References ........................................................................................................................ 87
Appendix Byggcement CEM II/A-LL 42.5 R, Product sheet
Sikament ECO 12-2, Product sheet
Sika UCS, Product sheet
1
1. Introduction
1.1. Background
Today the concrete industry is using a variety of additives and filler materials to get desired
properties of concrete. One type that’s still in research mode is cellulose fibres or fibrils. In
2010 a pilot study, conducted at Swedish Cement and Concrete Research Institute, CBI,
showed the capacity for pulp, micro cellulose fibres, from the forest industry to act as a
Viscosity Modifying Agent (VMA) in concrete. This project however was too limited to find
answers for optimal use of this kind of material. Now the forest industry company Stora Enso
wants to find out if their pulp can be used in concrete in order to improve its properties
somehow.
Therefore this thesis is carried out at the Swedish Cement and Concrete Research Institute,
CBI, on behalf of the company Stora Enso. Stora Enso is a forest- and paper industry
company in Sweden. Out of pulp used in paper production, Stora Enso has developed
microfibrillar cellulose for the purpose of mixing them into concrete. In today’s technical
society paper is in less demand, so therefore Stora Enso are investigating new markets for
their products [1].
With this work Stora Enso wants to find out if their microfibrillar cellulose fibres can act as
reinforcement in concrete. There are previous studies showing that the microfibrillar cellulose
act as a reinforcement in concrete, and as a cracking reducer and make the concrete more
sustainable [2, 3].
This thesis is a continuation and built on a previous literature survey made at Stora Enso. It is
also a continuation of a smaller earlier project on CBI, on behalf of Stora Enso, regarding
these fibrils. In the literature survey by Stora Enso they come to the conclusion that a good
dosage of MFC in concrete is about 3% microfibrillar cellulose of the cement weight. Our
dosages are smaller than that, but higher than the dosages used in the earlier project
performed at CBI.
1.2. Aim
This thesis aims at answering the question if microfibrillar cellulose from Stora Enso can act
as an additive in concrete, and thus, investigate if they can improve the concrete properties
either in its fresh state or in its hardened state. Tests will therefore be carried out to measure
the fresh and hardened properties of mortars containing microfibrillar cellulose.
To determine the properties in its fresh state, mortar rheology and workability will be
evaluated. Anti-wash out resistance will also be tested. In its hardened state flexural strength,
compressive strength, free shrinkage and ring cracking will be tested. A possible use of fibril
suspension as a surface cover on hardened concrete will also be looked at.
1.3. Research significance
Not many studies have been performed in the field of cellulose fibrils in concrete. Of the few
that has been done they often involve self compacting concrete [2, 3, 4, 5, 6]. It is an
2
important subject to cover, in the way that Sweden is rich in forest assets and therefore it is a
large part of the country’s economy. When the demand of paper decreases, paper companies
must find new markets for their products.
1.4. Methodology
The fresh properties of the mortars will be tested by mini slump test where the ability to flow
freely is tested. A viscometer will also be used in its fresh state. The viscometer used to
perform measurements of rheology, such as determining the parameters plastic viscosity and
yield stress according to the Bingham model (se section 2.2.1.1.). This is to ensure that the
fresh mortars have good characteristics in terms of fresh properties. Anti wash-out resistance
is measured by conducting an anti wash-out test. The hardened properties of mortar are tested
by pouring the mortar in a total of 12 prisms and a steel ring form. The sizes of the prisms are
160×40×40 mm, and the ring 250, 170 and 40 mm, outer diameter, inner diameter and height
respectively.
For the flexural and compressive strength measurements, nine of these prisms are used, three
prisms respectively for tests at 1, 7 and 28 days of age [7]. The last three prisms are for free
shrinkage tests which is measured at 1, 7, 14, 21, 28, 35, 63, 119 and 231 days of age [8]
including the day when each mortar cracks in the ring test. The reason for why three prisms
are tested every time is to secure the significance of the results and minimize measurement
errors. Worth noting here is the fact that due to the high required dosage of superplasticizer
(SP) for some mortar mixtures, the cement hydration for some mixtures were delayed to such
an extent that the specimens were impossible to demould the day after casting. These
specimens were instead demoulded one day later (one type were actually demoulded after
three days). The ring test is performed to see at what age the ring cracks and compare it with
the free shrinkage tests. The ring might crack during a weekend, and if so, it will be a span of
1-3 days during which the ring could have ruptured. There is nothing to do about this fact
except to deal with it, if this happens.
1.5. Limitations
The tested fibrils, originating from pine pulp, are in two sizes. The fine material named MFC1
has undergone more homogenization than the course material named MFC2. Two different
water-to-cement ratios (w/c) have been used, 0.45 and 0.60. One type of cement is used,
ordinary Portland cement, CEM II, A-LL 42.5 R.
Only concrete equivalent mortar have been tested (se section 4.1 for further information).
Four regular series have been tested, each with three different dosages of fibrils, 1, 2 and 3
kg/m³. Series one and two have a w/c of 0.60, while series three and four have a w/c of 0.45.
In series 1 and 2 the three fibril dosages of 1, 2 and 3 kg/m3 corresponds with 0.14, 0.27 and
0.41 percent fibrils of the cement weight. In series 3 and 4 these percentages are instead 0.11,
0.22, and 0.33 of the cement weight. The reason for the different percentages between the two
w/c is because the w/c 0.45 have a higher cement mass than the 0.60 w/c mortars. Two
reference mortars have been moulded, one for each w/c. Series one and three include the fine
fibril type, MFC1, while series two and four include the coarse fibril type, MFC2.
3
A series 5, containing four different mortars, one reference and three types each with fibrils as
mentioned before and fibril type MFC1, have also been moulded. These are only tried for
rheology and for flexural-, and compressive strength after 28 days. These specimens contain
the same amount of superplasticizer to eliminate a possible effect by different dosages of SP
in previous series.
The fibrils are also evaluated to see if they can be used as an anti wash-out admixture.
Last but not least the fibril suspension have been placed on the surface of a mortar to see if the
fibrils may act as a surface coverage.
Table 1-1: The principle test matrix.
Mix/Series W/C- MFC dosage MFC dosage MFC type Superplasticizer
ratio (kg/m³) (% of the
dosage
(-)
cement
weight)
Ref 0.60 0.60 0 0 N/A
Varies
Series 1
0.60 1 0.14 MFC 1
0.60 2 0.27 MFC 1
0.60 3 0.41 MFC 1
Series 2
0.60 1 0.14 MFC 2
0.60 2 0.27 MFC 2
0.60 3 0.41 MFC 2
Ref 0.45 0.45 0 0 N/A
Series 3
0.45 1 0.11 MFC 1
0.45 2 0.22 MFC 1
0.45 3 0.33 MFC 1
Series 4
0.45 1 0.11 MFC 2
0.45 2 0.22 MFC 2
0.45 3 0.33 MFC 2
Ref series 5 0.45 0 0 N/A
Constant Series 5
0.45 1 0.11 MFC 1
0.45 2 0.22 MFC 1
0.45 3 0.33 MFC 1
1.6. About Swedish Cement and Concrete Research Institute, CBI
This thesis is carried out at Swedish Cement and Concrete Research Institute, CBI, located in
Stockholm, Sweden. CBI is a publicly funded institution that is divided into five groups
consisting of research, development, testing & inspection, consulting & investigations and
courses & information in concrete & rock area. CBI has its facilities in Stockholm (head-
office), Borås and Lund. CBI's objectives are formulated: "CBI is our country's foremost
4
environmental R & D in concrete and rock areas and one of Europe's three leading industrial
research institutes in the same areas” [9].
CBI has the largest single engineering resource in the field of concrete, with researchers and
experts in the construction field. The Institute is also known internationally for being strong in
the field of aggregates and natural stone. On behalf of SP, the Technical Research Institute of
Sweden, CBI carries out testing of products and inspection of production control in several
areas, such as fresh concrete properties, aggregates, additives, filler materials, pigments for
concrete, asphalt, etc. Testing and certification are carried out on products that qualify for the
P-mark (SP's quality mark) [9].
The company's premises in Stockholm are equipped for a variety of controls such as for
example rheology, compressive strength, flexural strength, shrinkage and electron scanning
microscope to zoom in at the micro structure of the concrete.
5
2. Literature survey
2.1. Concrete
Concrete is the world’s most used construction material. It has good durability, formability
and strength [10]. It has a wide range of uses that ranges from small structures such as
fountains and sculptures to large scale complex industrial productions. Concrete is almost
always used in foundations, in building structures, bridges, tunnels, dams, roads, ports,
landing fields,...etc [11]. Concrete, as a construction material, is therefore a big part of the
modern building environment. Concrete, as a building material, has long history and dates
back to several hundred years B.C. We know with certainty that the Greeks and the Etruscans
produced a material much similar to the concrete in our time [10]. The concrete was used in
large amounts of structures for example houses, bridges, ports and aqueducts [5]. There are
still numerous of these structures intact; an example of this is the Pantheon in Rome [11].
The main constituents in concrete today are cement, water and aggregates. The most
commonly used cement type is Portland Cement which was discovered in the middle of the
19th
century by a man named Joseph Aspdin in Great Britain [10]. The aggregates most often
consist of stones and gravel from nature, either in the form of crushed gravel or as naturally
formed gravel.
Cement together with water is called the paste phase of concrete and while it hydrates, it binds
together the aggregates. The aggregates are often referred to as the skeleton of the concrete.
Chemical and mineral additives in concrete are also often used. Examples of chemical
additives are superplasticizers which disperse the particles and make the concrete more prone
to flow. It can also act as a water reducer. Accelerators are used to increase the hardening of
the concrete, and retarders to do the opposite. Air entraining agents are used to increase the
frost resistance of the hardened concrete [10, 12].
Mineral materials acting as an additive can vary from lime stone filler to silica fume. Steel
fibres are used to increase the concrete mechanical properties of the concrete [12].
By use of different additives there are numerous ways to vary the concrete’s properties, both
in the fresh and in the hardened state [10].
2.2. Fresh concrete properties
2.2.1. Rheology
The definition of rheology is “the science of flow and deformation of matter”[13]. Rheology
is used as a science in many fields, such as the paper industry and the food industry, in the
concrete field it is used to measure the properties of the fresh concrete. When categorizing
rheology the relationship between force, deformation and time is studied. With these
parameters it is possible to measure quantities such as viscosity and the yield stress [13, 14].
To describe the basic concept of rheology it can be explained by the parallel plate model. In
this model, the space between the two parallel plates is filled with a liquid. A force is applied
to the top plate and it will lead to a deformation of the liquid. The shear stress is defined as the
added force divided by the area of the plates. The deformation is called shear strain and can
be expressed by an angle γ, or as dx/dy. The angle is the angle between the lower and the
upper planes. The time it takes before the deformation γ is obtained is called shear rate (s-1
)
[13, 15].
6
Figure 2-1: Parallel plate model
For the fresh concrete to be able to flow, a certain force must be applied to the mass. When
the force increases, the shear stress (τ) increases, and when reaching a certain stress, the yield
stress (τ˳) and the rate of shear ( ) of the concrete can be measured. When combining the
torque and deformation of the concrete in a viscometer, the yield stress (τ˳) and the plastic
viscosity (μ) can be obtained via a linear regression and then applying a mathematical
equation called Reiner-Rivlin equation.
Figure 2-2. Model describing the shear stress (τ) and shear rate ( )correlation,
with factors yield stress (τ˳) and plastic viscosity (μpl).
When the water amount is lowered, and the w/c is constant, the plastic viscosity will increase,
whereas if more cement is added simultaneously, the plastic viscosity will only be changed
slightly. In conclusion the water amount has more effect on the concrete mix than the cement
amount has. If air is added to the mix the plastic viscosity will decrease. The yield stress is
higher the stiffer the mix is (absence of a superplasticizer, small amount of water or if it
contains lots of fines) [13].
7
The rheology of a suspension is a function of the rheology of the suspending media. Concrete
is often described in three scales, micro mortar, mortar and concrete.
In micro mortar suspensions the suspending media is water in which cement, filler and the
aggregates 0.125 mm are suspended. But in mortar the suspending media is micro mortar
where the sand fraction particles are suspended. Finally in concrete the mortar becomes the
suspending media in which gravel particles are suspended.
The conclusion is that in order to optimize the rheology of concrete, the small scale
suspensions need to be optimized [16].
Figure 2-3. Three scales of particle suspensions in concrete
The rheological properties of cement paste depend on the type of cement, specific surface area
of cement, water cement ratio, mixing procedure, time after mixing and temperature of
hydration [11].
A liquid can be defined as a Newtonian liquid, a time independent non-Newtonian liquid, a
non-Newtonian time dependant liquid or as a visco-elastic liquid.
Here we will only describe Newtonian liquids and non-Newtonian time independent liquids.
A Newtonian liquid shows a linear relation between the shear stress and rate of shear (figure
2-4). For these liquids the viscosity remains constant for all rates of shear. Liquids with this
flow behaviour are simple, single phase liquids and solutions of liquids with low molecular
weight. Examples of these liquids are water, oil, petrol and glycerol [13].
A Non-Newtonian liquid varies in viscosity with the rate of shear. Non-Newtonian time
independent liquids can be categorized as shear thinning, shear thickening, with or without a
yield stress. The behaviour of a shear thinning liquid is described by a decrease in viscosity
when the shear rate is increased. In other words the fluid becomes more easily workable with
8
increase rate of rate. Shear thickening fluids shows opposite behaviour, the viscosity increase
with increased rate of shear (figure 2-4) [13].
Liquids with a yield stress behave like a solid below the yield stress and like a liquid above
the yield stress [13]. These are called visco-plastic materials and can be described by two
different models. First the Herschel-Bulkley model where the liquid exhibit either a shear
thinning or a shear thickening behaviour together with a yield stress. The second type is called
the Bingham model whereas the liquid show a Newtonian behaviour after the yield stress is
reached (figure 2-4) [13]. Examples of Bingham liquids are cement paste, mortar and concrete
[13].
Figure 2-4. Different types of flow curves
Figure 2-5 explains how the workability of the concrete is related to the yield stress and
plastic viscosity.
Figure 2-5: Diagram showing yield stress (τ0) contra plastic viscosity (μ)
9
2.2.2. Workability
The properties of fresh concrete heavily depend on the type and amount of the constituent
materials of the concrete. To cite O.H. Wallevik on what constitutes a good concrete; “It is a
concrete that satisfies certain requirements with regard to strength, durability, and volume
stability, along with appearance, at the lowest possible cost” [13].
Everything that’s added to the mix and its quantity has an effect on the concrete properties.
Some have more effect than others, and some have less, depending on the ingoing parts. [10]
It is important that the concrete is homogeneous and have the same properties in all parts;
therefore it is important to avoid separation in fresh concrete [11]. Whether the concrete is
good or not also depends on how the concrete will be cast, if it will be pumped, cast into walls
or simply just cast in horizontal forms [11].
The workability is therefore important as to how the concrete will be applied.
The slump test is used to measure how well the concrete deforms using a simple cone and
gravity [13]. Adding water to the mix always leads to a more fluid concrete and both a lower
plastic viscosity and yield stress. On the other hand, by reducing the water in concrete the
plastic viscosity and the yield stress increases making the concrete stiffer. More water in the
concrete also reduces strength, delay the hardening and decrease the durability. Therefore
decreasing water in the mix increases the strength of the concrete. On the other hand, water is
important for the chemical processes needed for the hardening, so the amount of water in
concrete cannot be lowered too much [13].
Slump is the deformation of the mortar, while the slump flow however is how much the
mortar will flow outwards on the surface it is tested on, when the cone is lifted. If the slump
flow is held constant and the water is reduced and at the same time the superplasticizer is
increased to hold the slump flow constant, the plastic viscosity (μ) is increased while the yield
stress (τ˳) remains almost the same. During the lifespan of the fresh concrete, the yield stress
(τ˳) always increases, the plastic viscosity though depends on the amount of superplasticizer
in the mix. The shape of the aggregates also affects the plastic viscosity and the yield stress.
Round aggregates will result in a lower plastic viscosity than crushed aggregates, the same
results appear with the yield stress where the round aggregates will result in a lower yield
stress than if crushed aggregates are used [13].
2.3. Hardened concrete properties
Concrete can be produced with numerous different proportions of constituent materials. The
material can therefore have as many differences in properties as there are different ways to
proportion the concrete. Concrete can vary in its w/c, have different additives such as fly ash,
air entrainers, superplasticizers, steel fibres etc. to alter its properties.
There are some common properties though, for all types of concrete. The most typical
properties of hardened concrete are that its tensile strength is about a tenth of the compressive
strength. Because of its low tensile strength the concrete always leads to cracking, even at
very moderate loads. Most load bearing structures require reinforcement, often made up of
steel bars. The function of the steel bars is to transfer the tensile forces instead of the concrete
structure after the initial cracks have appeared [17].
10
The way the bars are anchored and the bond in the interface are important to the function of
the structure and have a direct affect on the crack widths. Concrete changes its volume during
its lifetime. The concrete shrinks gradually after it hardens, and continues to do so for a long
period of time. There are two types of shrinkage; drying and autogeneous shrinkage. The
autogeneous drying continues until no chemical process with water can occur any longer,
while the dehydration stops when no free water is left in the concrete. The drying conditions
are therefore important. The faster the dehydration occurs in the construction the faster and
bigger the cracks will appear. One other important factor is the creep of concrete. Creep
occurs gradually in concrete and is the result of long time load on the structure. Concrete is a
durable material. It is moisture resistant and can be made completely water proof. The
material doesn’t become mouldy, it isn’t burnable, it’s resistant to relatively high
temperatures, soundproofing and is heat storing [17].
2.4. Fibres in concrete
Previous studies have been done on cellulose fibres in concrete. One study that focuses on
fibres in general as reinforcement in concrete is carried out by Zollo [2]. In this report the
diameter of the fibres are in focus as well as the length-diameter aspect ratio. The diameter, in
the report by Zollo, range from 0.4 to 16 μm, and an aspect ratio varying from 40 to 1000. It
was found that the fibre efficiency depended heavily on the volume percentage of fibres in
relation to the concrete surface and the surface area of fibres and cross-sectional area off the
fibres across a given plane. Dense packing of the fibres was found to provide better durability
of the matrix. Finer fibres resulted in a denser packing of the fibres than the coarser fibre did,
and it showed that the matrix was stronger and reduced the propagation of cracks. The fibres
had an effect on the fracture energy by bridging propagation cracks. Therefore, it appeared
that the fibre strength and elastic properties are important for the initial crack propagation.
Jorge et al [18] has studied the compatibility of wood and cement. The conclusions of these
experiments which were carried out with three-layer panels made with sawdust and a core
with flakes from hard tropical wood is that the flakes length, thickness, and density have a
strong connection with the modulus of rupture, elasticity, water absorption and thickness
swelling. Thus, longer and thinner flakes resulting in stronger, more stable and stiffer panels
Vac Vicar et al. [4] have done a study on aging of commercially produced cellulose fibres
used as reinforcement in cement. The study shows a reduction in depot porosity and water
absorption in the aged products of cellulose fibre-reinforced composite cement products. The
matrix became denser and the interaction between the fibres and the matrix was improved
resulting in 46% increase in compressive strength after a 5-year period compared with the
unaged matrix. The same results were shown of the aged matrix when exposed to weather
conditions and accelerated aging.
Chun et al [3] have looked into the possibility of using recycled and repulped fibres (residuals
of the pulp and paper industry) in concrete, and the conclusion is that it improves durability
and that recycled fibres can be used in concrete to make it more sustainable, and the easier it
is to repulp the fibres, the better the cement appears to get considering durability.
Steel fibres are widely known for reinforcing the concrete in terms of increasing toughness
and energy absorption capacity, reduce cracking, and improve the flexural strength and the
11
durability of the concrete. Polymeric fibres can also be used, but not with the same results as
the steel fibres. Cellulose fibres though do not suffer corrosion problems as steel fibres nor do
they suffer from alkaline attacks unlike glass fibres. Cellulose fibres can be considered as an
alternative to polymeric fibres in the making of lightweight concrete and a more inexpensive
one too [5]. According to this report incorporating cellulose fibres into self consolidating
concrete in suitable proportions the fibres can reduce plastic shrinkage, bridge cracks and
delay their propagation. They may as well increase the mechanical resistance and act as a
rheological additive and as light weight filler. In this study it could also be concluded that
incorporation of cellulose micro fibres in the concrete decreases the need for superplasticizers,
and reduces the density of the concrete [5].
According to Zhengwu Jiang et al [6], when cellulose fibres were added to self compacting
concrete with high-volume mineral admixtures (HMSCC) the results showed an improvement
in impact resistance gradually with increased fibre dosage, but did little or nothing to the
compressive strength.
Also pointed out is that the properties of the fibres have a part of the properties of the
concrete, where fibre concentration, fibre geometry, fibre orientation and fibre distribution
play a part in the overall behaviour of the concrete. The surface of the concrete is also said to
positively be modified by improving the abrasion resistance.
2.5. Underwater concrete
Underwater casting has been used for quite some time in the industry. To achieve good results
when casting under water, the composition of the concrete is of importance, and so are the
workability and the method of casting. Hardening of the concrete is often good. However the
durability of underwater casted structures is often of varying quality, and this is often the
cause of bad casting. By the use of anti washout admixtures (AWA) a more even durable
construction can be received.
Conventional concrete is most often not compatible with AWA, bad resistance to frost is often
a problem, and this is because the air pore system is negatively affected by the AWA. The
result is large and badly distributed pores in the hardened concrete [19].
Progress in the field of underwater concrete has moved forward by incorporating new
methods of placing the concrete, inventing new chemical admixtures for concrete and by
implementing new cementitious materials to the concrete. An example of this would be in the
repair of hydraulic canals.
Flowable concrete has been used to repair hydraulic canals and floors of stilling basins that is
damaged by abrasion-erosion. These castings require that the concrete is easily placeable and
can move around obstacles easily, but yet maintain minimum dilution by water. The concrete
to be cast under water should have good rheological and mechanical properties that have a
direct affect on the performance. The composition and rheology of the concrete dictates the
resistance to be diluted by water and segregation [20].
As mentioned, the addition of an AWA can minimize the dilution of cement paste by water,
eliminate external bleeding and reduce the risk of segregation and sedimentation. According
to M. Sonebi and K.H. Khayat [20] the addition of a high-range water-reducing agent
(HRWR) combined with an AWA can furthermore stabilize the fresh underwater concrete
making it flowable but sufficiently viscous. In a report by Khayat [21] it is said that the use of
AWA reduces the amount of free water available for the lubrication of the concrete, and it is
12
therefore a necessity for HRWR to be incorporated. Another factor affecting the stability of
underwater concrete is the composition of the binder and the water-cement ratio (w/c).
The Standards of Japan Society of Civil Engineers (JSCE) have recommendations of limiting
the w/c up to 0.50 for underwater structures in sea water, and 0.55 in fresh water. Lower w/c
are recommended for structures that require more durability, such as marine piles which have
a recommended w/c of 0.38 – 0.42 [21].
13
3. Methods and materials
3.1. Equipment
3.1.1. Laboratory
All procedures and tests were performed in a laboratory environment having a temperature of
20±2 ˚C (according to SS-EN 196-1:2005 [7]).
The specimens tested for free shrinkage and the shrinkage rings, were placed in a room with a
temperature of 20 ˚C and a humidity of 50%, where the measurements also took place. The
other specimens, tested for flexural strength and compressive strength were placed in a room
with a temperature of 20 ˚C and a humidity of 100% until the day the strength tests were
performed.
Figure 3-1: Picture of one of the labs
3.1.2. Sieves
Two kinds of sieves have been used; one coarse to remove all aggregates over 4 mm (figure
3-2 left) and one fine series of sieves (figure 3-2 right), according to SS-EN 196-1:2005 [7],
in order to get a proper analysis of the particle size distribution of the aggregates used for the
mortars.
For the sieve analysis, an amount of 5 kg aggregates were weighted and divided 3 times in a
divider for aggregates (figure 3-3), using approximately 0,625 kg off aggregates in the sieve
according to SS-EN- 932-2 [8].
Three batches all together were sieved in the coarser type; each containing about 330 kg of
aggregates, and each of these three batches were then sieved two times in the finer sieve for a
proper particle size distribution curve. Together there are six distribution curves (fig 3-15),
two curves for each batch. A mean is calculated for every time a new batch is sieved, resulting
in three different distribution curves which we used and based our mortar recipes on. To
clarify, for each recipe we’ve mixed the actual distribution curve for that batch of aggregates
have been taken into account when proportioning the mortar recipe.
14
Figure 3-2: The two different sieves, the course one (left), and the finer type (right).
Figure 3-3: Divider for aggregates
3.1.3. Mixer
The mortars were mixed in a Hobart Food Mixer (figure 3-4). The mixer consists of a
stainless steel bowl with a capacity of 5-6 liters that is fixed to a mixer frame. Attached to the
mixer frame is a stainless steel paddle that revolves about its own axis in a planetary motion
that is controlled by an electric motor. The mixer works at three different speeds [7].
Figure 3-4: The Hobart mixer
15
3.1.4. Air meter
For the measurements of pore volume an air meter, shown in figure 3-5, were used.
The basic concept is to fill the apparatus with mortar, fill up the remaining volume of the
meter with water so no air is left in the apparatus except for the air in the pores of the mortar.
Then a pressure is built up mechanically with a pump. The pressure, to compress the mortar,
is then released and air fills up the empty space in the apparatus made by the compression of
the mortar. The air volume that fills up the apparatus is the air content of the mortar.
Figure 3-5: The air meter
3.1.5. Concrete viscometer
A concrete viscometer is an instrument used to determine the rheology of fresh cementitious
materials, like paste, mortar and concrete, i.e. the material viscosity and yield stress [13]. The
viscometer used in this project is a ConTec 4 (figure 3-6).
The apparatus consists of a rotating outer cylinder and a stationary inner cylinder that both are
concentric. The inner cylinder will be lowered into the outer cylinder where the fresh mortar
is placed [15]. The measurement is done by introducing a shear in the concrete with the
rotating outer cylinder and then measuring the torque with the inner cylinder [13].
The inner cylinder measure the torque whiles the outer cylinder rotates and measure rotational
speed and together they are connected to a computer, which convert these parameters to shear
stress and shear rate. With help of the Bingham model, rheological parameters such as yield
stress and plastic viscosity is calculated. All calculations are done with a program called
FreshWin 4.0.
The test is carried out two times for each mix and then an average value is taken from the two
tests regarding the yield stress and the plastic viscosity [13].
16
Figure 3-6: The ConTec 4 viscometer.
3.1.6. Flexural strength testing machine
For testing of flexural strength the three point loading method is used that meets the
requirements according to SS-EN 196-1:2005 [7]. The machine used is a MTS 500 (figure 3-
7).
Figure 3-7: The flexural strength testing apparatus
17
3.1.7. Compressive strength testing machine
The machine for testing compressive strength is a MTS 4500 and meets the requirements
according to SS-EN 196-1:2005 [7] (figure 3-8).
Figure 3-8: The compressive strength testing apparatus
3.1.8. Free shrinkage testing apparatus
For the measuring of free shrinkage, an apparatus called Mitutoyo Absolute is used (figure 3-
9).
Figure 3-9: The Mitutoyo free shrinkage testing apparatus
18
3.1.9. Capillary porosity
When measuring capillary porosity a balance capable of weighing a test specimen to an
accuracy of ± 0.1 % of the mass of the specimen is required, along with a timer accurate to at
least one second. The specimens is placed in a water tank with height, width and depth of 600,
400 and 100 mm respectively. In the bottom of the tank a plastic mat with points placed to
keep the specimens at least 5 mm from the bottom of the base. This equipment meets the
requirements according to SS-EN ISO 15148 [22].
Figure 3-10. Specimens for series 1 and 2 in the water tank,
and the plastic mat with supporting points.
3.1.10. Moulds
The moulds for the prisms have all the same dimensions, 160x40x40 mm. The moulds for
prisms intended for flexural- and compressive strength testing are made of steel (figure 3-11).
The moulds for the prisms used to measure free shrinkage (figure 3-12) are instead made of
Styrofoam to be able to supply the prisms with a metal stud needed for the free shrinkage
measurement procedure (figure 3-12). This is for making the measurement of the specimens
more reliable. The constrained shrinkage moulds are made of wood and steel (figure 3-14).
The dimensions are 400 mm of height, outer dimension 250 mm and inner dimension 170
mm. A steel ring placed in the middle of the mould is what prevents the mortar from
shrinking freely. The moulds for the specimens intended for capillary porosity have the
dimension 100x100x100 mm (figure 3-13), and is made of steel.
19
Figure 3-11: Moulds for prisms for flexural- and compressive strength testing, empty (left) and filled
with mortar (right).
Figure 3-12: Moulds for free shrinkage without (left) and with mortar (right).
Figure 3-13: The metal stud at the end of the free shrinkage specimen, and a non used metal stud on
top of the specimen (left), the mould for the water capillary tests (right)
20
Figure 3-14: Mould for the constrained shrinkage (left), and shrinkage ring with mortar (right).
21
3.2. Materials
3.2.1. Aggregates
The aggregates consists of sand with a maximum size of 4 mm, mostly made up of quartz and
feldspar. The aggregates can be described as the skeleton of the mortar. In ordinary concrete
structures aggregates up to 32 mm in size can be used, but in our mortars we use a maximum
size of 4 mm. The density of the aggregates is 2.65 kg/m3. In order to remove all aggregates
over 4 mm from our original gravel the mechanical sieve, described in section 3.1.2, has been
used. Then a fine series of sieves were used, also described in section 3.1.2, to get proper
particle size distribution curves, seen in figure 3-15.
Figure 3-15: A batch of 4-mm aggregates
Figure 3-16. Particle size distribution curves
0
10
20
30
40
50
60
70
80
90
100
<0,063 0,063 0,125 0,25 0,5 1 2 4 8 16 31,5 63
Pass
ing a
mou
nt
[%]
Sieve size [mm]
Sieve 1
Sieve 2
Sieve 3
Sieve 4
Sieve 5
Sieve 6
22
3.2.2. Cement
Cement is a hydraulic binder in powder form and when mixed with water it hydrates and
forms into a hard mass. Cement affects many properties of concrete. Its chemical composition
is important in the fresh concrete's manageability and consistency, and the durability and
strength of stiff concrete.
The cement used in this work is Portland-limestone cement, CEM II/A-LL 42.5 R [appendix].
The Portland-limestone cement satisfies the requirements according to SS-EN 197-1 [24] and
is mainly used in residential structures in Sweden. It has a Blaine value of 430 m²/kg and a
compact density of 3080 kg/m³. The cement consists of Portland clinker, gypsum and finely
ground limestone. It’s produced in Skövde and Slite, Sweden. The cement is delivered by the
Swedish cement supplier Cementa AB.
Figure 3-17: Bag of cement
3.2.3. Fibrils
The tested fibrils, originating from pine kraft pulp, have by Stora Enso both been chemically
and mechanically treated, to release the fibrils [1].
The fine material, MFC1, has undergone more homogenization than the course material,
MFC2. Both were delivered as suspensions with solid content of 3% and are quite viscous,
which make them harder to apply to the mixes. Therefore both suspensions have been diluted
so the solid content has been approximately 1.5%. This has been done with distilled water, to
avoid contamination and risk of worsening the properties of the fibrils. The dilution have been
done with a simple kitchen blender (figure 3-18), mixing a 1/2 part of the 3% suspension with
1/2 part of distilled water. The mixing procedure has been 2 minutes for each litre of new
suspension.
23
Figure 3-18: The kitchen blender used for mixing the suspensions.
3.2.4. Superplasticizer
Superplasticizers are mainly used to give the concrete a more fluid consistency. It can also be
used to give the same consistency with lower water content and thus at a lower water-cement
ratio. With superplasticizers the water content in the concrete can be reduced by 10-30%
compared to a mixture without superplasticizers. If used like this, it will increase the strength
of the concrete and reduce shrinkage [11].
Superplasticizers have very little or no retarder capacity, in appropriate dosages. It is possible
to change the consistency of concrete from plastic to full flowing when recommended amount
of superplasticizer is used [11].
The agent Sikament ECO 12-2 is the SP we’ve used in our mortars. It improves the concrete
workability and give concrete a higher strength with time. Sikament ECO 12-2 works with
any type of concrete qualities and leads to fill out the matrix easier and contributes to higher
short term strength. The medium should be administered no earlier than when the water is
added to the mixture and excessive use can lead to stone separation. Dry content of Sikament
12-2 is about 17% [appendix].
Figure 3-19. The SP (superplasticizer) used in our mortars.
24
3.2.5. Anti-washout admixture
Anti washout admixtures is used to reduce washout of finer particles such as cementitious
materials and sand from fresh concrete when it’s placed under water. When an anti washout
admixture is mixed with concrete the viscosity of the concrete increases and the resistance
against segregation is improved. Anti washout admixtures in concrete prohibits the concrete
from bleeding, which means that the aggregates and the cement paste are kept homogenous in
the mix. Anti washout admixtures are used in all types of concrete structures which are
moulded in underwater environments [11, 19].
In this thesis one type of Anti washout admixture called Sika UCS were used.
Sika UCS is an anti washout admixture in powder form that is specially designed for
underwater concrete. It is compatible with all types of standard types of cement, included
Portland based cements. Sika UCS increases the self compacting properties of the concrete.
Some of the benefits of using the product are increased cohesion of the concrete, extended
processing time and minimizing of separation [appendix].
Figure 3-20: Anti wash-out admixture
3.2.6. Water
Regular tap water has been used in the mixing procedure of the mortars.
25
3.3. Methods
3.3.1. Rheology and workability
3.3.1.1. Mini slump flow
A mini slump flow is used to determine the free horizontal flow of the mortar, and to indicate
if the mortar is homogeneous. Homogeneous means that the water, paste and the aggregates
are intact and not separated from each other.
The test is carried out with a cone (figure 3-21), with height, top diameter and base diameter
of 50, 70 and 100 mm respectively. The cone is to be placed on a horizontal surface or disc.
The surface should be clean and the cone should be placed centrally on the disc. The cone is
filled to the top with the ready mixed mortar without vibration or shock. Then, the cone is
lifted vertically upwards until all mortar has been poured out of the cone on the disc or the
surface. The mortar will flow outwards on the surface and form a circular patty. This circular
mass will then be measured by two diameters perpendicular to each other and then calculate
the mean of these. The mortar is also examined to see whether it is homogeneous or not. This
will be evident if water has drained out at the edge of the circular mass or not, or if aggregates
are still left in the middle. The same procedure is done by Erdem et al [28] where they seek a
correlation between self-compacting concrete and concrete-equivalent mortar.
Our trials are set to have a slump flow between 200 – 250 mm. To receive a mini-slump flow
in this range, different dosages of the superplasticizer will be used.
Figure 3-21: The mini slump cone filled with mortar (left) and an example of spread mortar (right).
3.3.1.2. Air content
The measurement of air content of the mortars has been made by using an air meter (figure 3-
5). The procedure is as mentioned in section 3.1.4.
3.3.1.3. Rheology
The measurement of the rheology is done by using a machine called ConTec 4 (figure 3-6).
The procedure is as mentioned in section 3.1.5.
26
3.3.2. Mechanical properties
3.3.2.1. Flexural strength
The flexural strength of the specimens were measured by using the three point loading
method using a flexural strength testing machine [7]. The specimen was set centrically (figure
3-7) between the two lower steel rollers and the top steel roller. The top steel roller was added
a load of 50 N/s that developed a bending moment to the specimen. When the load exceeds
the ultimate load of the material, the specimen broke and a value for flexural strength can be
calculated. For a better measurement accuracy, the test was performed with three individual
specimens of the same mixture. The mean value was then calculated for the three specimens.
The testing method meets the requirements according to [7] SS-EN 196-1:2005.
3.3.2.2. Compressive strength
For measuring the compressive strength of the prism a compressive strength testing machine
was used (figure 3-8).
The specimen was placed between the plates of the machine to transmit the load of the
machine to the surface of the mortar specimen [7]. The top plate was added a load of 2400 N/s
that developed a compressive pressure to the specimen. When the load exceeds the ultimate
load of the material, the specimen broke and a value for compressive strength can be
calculated. For a better measurement accuracy, the test was performed with six different
specimen of the same mixture. The three specimens from flexural strength measurement were
divided into two pieces, which created six test specimens for compressive strength
measurements. The mean value was then selected for these six specimens. This testing
method meets the requirement according to SS-EN 196-1:2005 [7].
3.3.3. Shrinkage
3.3.3.1. Free shrinkage
The measurement of free shrinkage was made with a free shrinkage testing apparatus made
for mortars (figure 3-9). The method we used is based upon the SS-13-72-15 [26] standard
with the following deviations: the geometry is different, mortar instead of concrete is used and
the specimens were stored in a conditioning room of 50% humidity right after demoulding,
instead of storing the specimens seven days in an environment of 100 % humidity.
The method includes taking each of the three prisms made for this specific purpose and place
it in the apparatus, and measure the change in length. First a calibration is performed by
placing a piston that has the same original length as the prisms poses from the beginning after
the demoulding, between two fixated pistons. After that the deviation can be measured by
placing the prism between the two fixated pistons, see figure 3-9. This is done down to a
1/1000 of a millimetre. This is performed 3 times for each direction of the prisms, i.e. the
prism is flipped over to measure it from two directions. For each mixture three specimens are
tested and the mean of these three measurements are calculated.
27
3.3.3.2. Ring test
This test method is carried out to determine the time of cracking by using a steel form (figure
3-14) in which the mortar is restrained from free shrinkage according to ASTM standard C
1581-04 [27].
Deviations from this standard is as follows; (1) the inner steel ring has a height of 40 mm, and
outer diameter 170 mm, while the outer steel ring has a diameter of 250 mm, instead of the
measurements 150, 330 and 406 respectively, (2) no strain gages are used to electronically
monitor strain development or cracking in our specimens, (3) no data acquisition system are
used.
The test form that is used consists of a base of wood, an inner circular steel ring and an outer
ring of steel surrounded by wood.
The mortar is poured into the form between the two circular steel rings. A compressive strain
is build up in the steel ring caused by the restrained shrinkage of the mortar. Since the mortar
is prohibited from shrinking inwards by the steel ring it is forced to crack. The age of the
crack is registered simply by ocular observation.
The age of cracking of the rings and the measuring of free shrinkage of the prisms are
indicators of the materials resistance to cracking under restrained shrinkage.
Figure 3-22: Picture of mortar cracking while hindered from shrinking
3.3.3.3. Surface coverage
This test was performed to investigate if the microfibrillar cellulose as a surface cover could
prevent the mortars from shrinkage and cracking. Mortars with a w/c of 0.45 were cast on top
of six pre-made concrete tiles in three pairs. The three pairs were treated in 3 different ways,
no treatment of the surface, the surface covered with fibril suspension and the surface covered
with a plastic film. These three “treatments” a done so that they can easily be compared with
each other to see what happens when the concrete may shrink freely, the concrete is (maybe)
prevented from shrinking by the fibril suspension and lastly the concrete is prevented from
drying out.
The fibril type used in these tests was MFC1 with a solid concentration of 3%. The amount of
fibril that was added formed a layer of approximately 2 mm. To force the mortars to one side
drying, they were sealed on the sides and the edges of the mortars.
28
3.3.4. Wash-out test
This test method is performed to measure the amount of cement paste that washes out of a
sample of fresh mortar upon contact with water. The test method we use for this thesis is a
modification of the CRD C61 test adopted by the US Corps of Engineers [28]. A cylindrical
plastic bucket (figure 3-23 right) containing approximately 20 litre is filled with regular tap
water. A sample of freshly mixed mortar having a mass of approximately 2.0 kg is poured
into a container made of a steel mesh, with a maximum opening of 2 mm (figure 3-23 left). A
rope is attached to the top of the container so that it can easily be lowered into the bucket of
water. The container with the mortar will be lowered into the bucket and raised nine times.
Before the first immersion and after each immersion the remaining mass of concrete is to be
weighted to see how much of the mortar is washed out by the water. A total of ten
measurements are to be taken.
A total of five different recipes have been tested, all with a w/c of 0.45, and the fine fraction
of 100 – 400 μm fibrils. One mortar contained a regular Anti Washout Admixture (AWA),
with a dosage of 2 kg/m³ AWA, while the other four test mortars was a reference mortar
without any fibrils and three mortars containing three different dosages of fibril, 1, 2 and 3
kg/m³ fibril. The AWA-type is Sika UCS [27].
Figure 3-23: The receiving container empty (left), and after an immersion (right).
3.3.5. Capillary porosity
This test method is for determining by partial immersion in water the absorption coefficient in
each type of mortar [21]. The cubes moulded for each recipe are to be stored one week in a
conditioning room with a humidity of 100% and a temperature of 20 °C. After that they are to
be stored in an environment of 65% humidity and 20 °C for at least 21 more days. Then the
specimens are placed in a oven with a temperature of 40 °C until the specimens are stabilized
within 0,1% of its total mass, when measured over 24 hours, SS-EN ISO 15148 [21].
This test method consists of determining the change in mass from the first placement in the
tank and then weight the specimen each time it is taken out of the tank for a maximum of 24
h. Our test specimens have the dimensions of 100×100×30 mm, which has been sawed out of
a cube moulded with each recipe. The sides of the specimen shall be sealed with a watertight
tape to not allow water to enter the pores from the sides. A tank is filled with water and the
29
test specimen is placed on the bottom of a tank. The specimens shall rest on point supports to
make as much of the area of the specimen is free from the bottom of the tank and make as
much area as possible available for the water to absorb. The water level shall be 5±2 mm
above the highest point on the base of the specimen. Weighting of the test specimens shall be
made both before and after it is immersed into the water. The specimen is to be weighted after
5 min, 20 min, 1 h, 2 h, 4 h, 8 h and 24 h after immersion. Before weighing of the specimens
the surfaces should be blotted with a damp sponge to remove all free water on the surfaces.
If damp spots appear on top of the specimens, the procedure is to be terminated and the
specimen is to be considered water saturated.
Figure 3-24: A sawed specimen in series 2, to be tested for capillary suction, and one sawed
with water proof tape.
The tests were performed according to the SS-EN ISO 15148 [21] standard with a few
deviations. Instead of three specimens for each type, only one were tested. This was because
of the limitation of time and quantities of moulded mortar. The standard says that after 8 h a
minimum of two more weightings should be performed including the 24 h weighing. Due to 8
hour working day and no chance of staying in the lab after hours, no additional weighing
occurred between the 8 h weighing and the 24 h weighing. That means in a 16 h period no
weighing were performed, and the specimens which had absorbed so much water that the
upper surface had become blotted with damp spots in this period have no accurate time where
the tests should have been terminated, which the standard says they should be.
When the weightings are done the measured values are plotted in a graph. The difference in
the mass of each weighing and the starting mass are divided by the area of the specimen in
contact with water, Δmt =(mt - mi)/A . This is then plotted against the square root of time of
each weighing, √ .
If the graph shows a straight line this should then be extended to zero, where it cuts the
vertical axis, Δmʹ0.
The water absorption coefficient is then calculated from
Aw = (Δmʹtf - Δmʹ0) / √ f
which means that the slope of the line is the water absorption coefficient.
See section 5.6 for further explanation.
30
31
4. Mixture design and methodology
4.1. Concrete equivalent mortar
The basic idea behind concrete equivalent mortar, CEM, is to replace the coarse aggregate
fraction in the concrete by sand with an equal surface area. The total aggregate area to be
coated with paste is then the same for the mortar as it was with the coarse aggregates. The
reason for the development of this method was because of the complex design of self
compacting concrete (SCC) and the time, effort and resources it takes to mix materials and
investigate parameters of concretes such as SCC in the laboratory. With this method it takes
lesser time and resources to prepare the mortars, than it takes to make concrete mixtures.
When making large batches with many specimens it is very handy to take advantage of this
method. Studies have been made on the relationship of the rheological properties between
concrete equivalent mortars and SCC [28].
Shown in table 4-1 is the basic concrete recipes we’ve based our mortar recipes on. The two
CEM-recipes are our two reference mortars in series 1 to 4. When lowering the w/c from 0.60
to 0.45 the water amount is almost the same while the big change in the recipes is in the
cement amount where the lower w/c has a higher amount of cement. In figure 4-2 the basic
principle of QEM is presented graphically.
When proportioning our concrete equivalent mortars, we always base them on the distribution
curve of the current aggregates used for the particular mortar recipe.
Smaller aggregates have a larger area than bigger aggregates. Therefore a smaller volume of
fine aggregates is needed for the mortar, than if coarser aggregates are used with the same
specific area.
Table 4-1. Concrete recipes and corresponding Concrete Equivalent Mortar
Material Concrete (kg/m³) CEM (kg/m³)
W/C 0.60 W/C 0.45 W/C 0.60 W/C 0.45
Cement 320 430 738 929
Sand 0-4 mm - - 788 689
Water 192 193.5 443 418
Sand 0-8 mm 1088 943 - -
Coarse aggregate 8-16 mm 725 771 - -
CA% 40 45 - -
32
Figure 4-2: Schematic difference in volume of the constituents in concrete (left), and in concrete
equivalent mortar (right)
4.2. Mixing sequence
The following procedures have been used for mixing of the mortars. The mixing is carried out
by first weighing the aggregates, cement, water and eventual additives, including fibrils and a
superplasticizer, figure 4-3.
After that the ingredients were blended in a mixer (figure 3-4). First the cement and the
aggregates were mixed together for 1 minute at low speed, gear 1. Then the water was added
during the mixing. If there was fibrils included in the recipe the fibrils was mixed together at
the same time as the water. When superplasticizer was used, it was added at the same time as
the water and fibrils. After that a visual inspection is required including a manual stir of the
batch, since the mixing paddle doesn’t completely reach the bottom of the bowl. Then the
mortar was blended for 5 minutes at high speed, gear 2.
Figure 4-3: Scales for weighing the constituent materials for the mortars
33
4.3. Characterization of fresh mortar
When the mixing was completed, a mini slump flow test was carried out. The target was set
between 200 – 250 mm. After that the rheology was measured with the ConTec 4 viscometer
(figure 3-69.
Each batch was also controlled for air content, using an air meter (figure 3-5).
4.4. Moulding of the mortars
When the tests of the fresh mortars were done the mortars were moulded (see section 3-1-10).
For each batch three prisms are moulded for measuring free shrinkage, nine prisms are
moulded for the flexural- and compressive strength tests, one cube for capillary porosity, and
one ring is moulded to determine time of cracking.
4.5. After demoulding
The flexural and compressive strength tests are done 1, 7, and 28 days after moulding [7]. The
prisms for free shrinkage are measured every week beginning on the day the specimen are
demoulded, and thereafter on day 7, 14, 21, 35, 63, 119 and 231 after the moulding [26].
Since we started counting the days from the day the specimens were moulded, and not from
the day they were demoulded, as should, the measurement of the free shrinkage were not
performed on exactly the days the standard says they should be measured. Some specimens
couldn’t be demoulded until day 2 or day 3 because of a high dosage of superplasticizer
delaying the hardening. Therefore the days the testing was carried out varies between the
series and the mortars. The shrinkage rings are inspected every day from the day they are
moulded, except for on weekends.
The specimens tested for free shrinkage and the shrinkage rings, were placed in a room with a
temperature of 20 ˚C and a humidity of 50%, where the measurements also took place. The
other specimens tested for flexural strength and compressive strength were placed in a room
with a temperature of 20 ˚C and a humidity of 100% until the day the strength tests were
performed. This procedure isn’t accordingly to the SIS standard [26]. The deviations made in
the free shrinkage procedure are explained in section 5.4.
34
35
5. Results and discussion
5.1. Fresh mortar properties
The mortars are split into the five different test series. Series 1 and 2 and its associated
reference mortars have a w/c of 0.60. Series 3 and 4 and its reference have the w/c of 0.45,
and the 5th
series also have a w/c of 0.45 and its own reference mortar with the same w/c. The
results for series 1 and 2 are further on split into non-plasticized and plasticized. For series 3
and 4 no tests were performed on non-plasticized mortars for the reason that the reference
mortar was already stiffer than the target mini-slump flow of 200 mm (it had a mini-slump
flow of 108 mm).
It was from the start intended to do only four series. A 5th
series were decided for the sole
purpose of finding out what increasing the fibril dosage in the mortars do for the strength with
no other additives, like the superplasticizer, possibly inflicting on the build up of strength in
the mortars. This series was only tested for flexural and compressive strength after 28 days.
Air content was measured for all mortars that were moulded, with one decimal of accuracy.
When proportioning concrete the usual amount of air content included is about 2 %.
5.1.1. Series 1
The rheology and workability results for the non-plasticized mortars in the first series are
shown in table 5-1. It is evident that the addition of MFC1 influences the fresh properties
considerably. Without SP, the reference mortar reached a slump-flow of 275 mm with the
corresponding 19 Pa yield stress and 0.5 [Pa s] plastic viscosity. With the gradual increase of
MFC1 dosage, the mini slump-flow eventually decreases to the minimum slump flow of 100
mm. The yield stress increases considerably to 463 Pa while the plastic viscosity increases to
3.1 [Pa s] for the highest MFC1 dosage of 3 kg/m³. Note that 100 mm is the diameter of the
base of the slump-cone, so this is the lowest measurable value. The values in table 5-1 are
plotted in figure 5-1-1 (mini slump-flow), figure 5-1-2 (yield stress) and in figure 5-1-3
(plastic viscosity).
Table 5-1: Pre-test results for non-plasticized mortars in series 1
W/C
(-)
MFC
dosage
(kg/m³)
Mini
slump flow
(mm)
Yield stress
(Pa)
Plastic viscosity
(Pa s)
0.60 0 275 19 0.5
0.60 1.00 153 117 0.4
0.60 2.00 102 232 0.2
0.60 3.00 100 463 3.1
36
Figure 5-1-1: (Pre-test results) Mini slump flow for non-plasticized mortars in series 1
Figure 5-1-2: (Pre-test results) Yield stress for non-plasticized mortars in series 1
100
150
200
250
300
0,0 1,0 2,0 3,0
Min
i sl
um
p f
low
(m
m)
MFC dosage (kg/m³)
0
50
100
150
200
250
300
350
400
450
500
0,0 1,0 2,0 3,0
Yie
ld S
tres
s (P
a)
MFC dosage (kg/m³)
37
Figure 5-1-3: (Pre-test results) Plastic viscosity for non-plasticized mortars in series 1
A superplasticizer was added to the mortars to target a mini slump-flow of approximately
200-250 mm. The results are shown in table 5-2. Note the increase in SP dosage reaching up
to as high as over 13% for the mortar with 3 kg/m³ MFC1. The yield stress increases, as
expected, when adding the superplasticizer but the plastic viscosity only changes moderately,
or even get lower, as the case is with the mortar containing 1 kg/m³ MFC1. This is however
very little in plastic viscosity change, and has no real significance. For the 2- and 3 kg/m³
fibril mortars, the plastic viscosity increases 2 times and 3 times the value respectively of the
reference mortar. Still the change in plastic viscosity is very low. The values in table 5-2 are
plotted in figure 5-1-4 (mini slump-flow), figure 5-1-5 (yield stress) and in figure 5-1-6
(plastic viscosity).
Table 5-2: Test results for plasticized mortars in series 1
W/C
(-)
Plasticizer
dosage
% *
MFC
dosage
(kg/m³)
Mini
slump flow
(mm)
Yield
stress
(Pa)
Plastic
viscosity
(Pa s)
Air content
(%)
0.60 0 0 275 19 0.5 0.0
0.60 0.94 1.00 235 38 0.3 1.2
0.60 6.96 2.00 245 21 1.1 0.8
0.60 13.38 3.00 207 46 1.5 0.8
* Weight-% of cement
0
0,5
1
1,5
2
2,5
3
3,5
4
4,5
5
5,5
0,0 1,0 2,0 3,0
Pla
stic
Vis
cosi
ty (
Pa s
)
MFC dosage (kg/m³)
38
Figure 5-1-4: Mini slump flow for plasticized mortars in series 1
Figure 5-1-5: Yield stress for plasticized mortars in series 1
100
150
200
250
300
0,0 1,0 2,0 3,0
Min
i sl
um
p f
low
(m
m)
MFC dosage (kg/m³
0
50
100
150
200
250
300
350
400
450
500
0,0 1,0 2,0 3,0
Yie
ld S
tres
s (P
a)
MFC dosage (kg/m³
39
Figure 5-1-6: Plastic viscosity for plasticized mortars in series 1
5.1.2. Series 2
The results of the non-plasticized mortars in series 2 are shown in table 5-3. The mortars in
series 2 are the first recipes mixed in the beginning of the project, and to get a feel for the
fibrils we started out with low dosages of MFC2. Some of these recipes are presented in table
5-3 and figures 5-1-7, 5-1-8 and 5-1-9.
The same trend in this series as in series 1 is found with the coarse fibril type MFC2. The
lower plastic viscosity value for the mortar containing 2 kg/m³ MFC2 compared to the mortar
containing 1 kg/m³ MFC2 is somewhat dubious in that the higher MFC2 dosage should
theoretically result in a higher plastic viscosity. The values in table 5-3 are plotted in figure 5-
1-7 (mini slump-flow), figure 5-1-8 (yield stress) and in figure 5-1-9 (plastic viscosity).
0
0,5
1
1,5
2
2,5
3
3,5
4
4,5
5
5,5
0,0 1,0 2,0 3,0
Pla
stic
Vis
cosi
ty (
Pa s
)
MFC dosage (kg/m³
40
Table 5-3: Test results for non-plasticized mortars in series 2
W/C
(-)
MFC dosage
(kg/m³)
Mini slump flow
(mm)
Yield stress
(Pa)
Plastic viscosity
(Pa s)
0.60 0 270 19 0.5
0.60 0.42 233 32 0.6
0.60 0.64 216 61 0.7
0.60 0.85 205 60 0.8
0.60 1.00 194 70 0.9
0.60 2.00 123 196 0.6
0.60 3.00 100 382 3.4
Figure 5-1-7: (Pre-test results) Mini slump flow for non-plasticized mortars in series 2
100
150
200
250
300
0,0 1,0 2,0 3,0
Min
i sl
um
p f
low
(m
m)
MFC dosage (kg/m³)
41
Figure 5-1-8. (Pre-test results) Yield stress for non-plasticized mortars in series 2
Figure 5-1-9: (Pre-test results) Plastic viscosity for non-plasticized mortars in series 2
0
50
100
150
200
250
300
350
400
450
500
0,0 1,0 2,0 3,0
Yie
ld S
tres
s (P
a)
MFC dosage (kg/m³)
0
0,5
1
1,5
2
2,5
3
3,5
4
4,5
5
5,5
0,0 1,0 2,0 3,0
Pla
stic
Vis
cosi
ty (
Pa s
)
MFC dosage (kg/m³)
42
The superplasticizer was added to level the mini slump-flow in series 2 for all mortars to the
target interval of 200-250 mm. Note that now only the same MFC2 dosages as for series 1 are
included (1, 2 and 3 kg/m³). Again low values of the plastic viscosity are measured for the
mortars containing the MFC2 dosages of 1 and 2 kg/m³. The resulting workability and
rheology values are shown in Table 5-4 and visualized in figure 5-1-10 (mini slump-flow),
figure 5-1-11 (yield stress) and in figure 5-1-12 (plastic viscosity).
Table 5-4: Test results for plasticized mortars in series 2
W/C
(-)
Plasticizer
dosage
% *
MFC
dosage
(kg/m³)
Mini
slump flow
(mm)
Yield
stress
(Pa)
Plastic
viscosity
(Pa s)
Air content
(%)
0.60 0 0 270 19 0.5 0.0
0.60 0.54 1.00 228 51 0.3 1.0
0.60 2.01 2.00 250 22 0.4 1.4
0.60 4.82 3.00 237 21 1.3 0.6
* Weight-% of cement
Figure 5-1-10: Mini slump flow for plasticized mortars in series 2
100
150
200
250
300
0,0 1,0 2,0 3,0
Min
i sl
um
p f
low
(m
m)
MFC dosage (kg/m³)
43
Figure 5-1-11: Yield stress for plasticized mortars in series 2
Figure 5-1-12: Plastic viscosity for plasticized mortars in series 2
0
50
100
150
200
250
300
350
400
450
500
0,0 1,0 2,0 3,0
Yie
ld S
tres
s (P
a)
MFC dosage (kg/m³)
0
0,5
1
1,5
2
2,5
3
3,5
4
4,5
5
5,5
0,0 1,0 2,0 3,0
Pla
stic
Vis
cosi
ty (
Pa s
)
MFC dosage (kg/m³)
44
5.1.3. Series 3
For series 3 and 4 only results from plasticized mortars are shown. Each series have a prelude
of initial tests to find the right dosage of SP to get within the target interval of a 200-250 mm
mini-slump flow. And when trying to find the right dosage for the mortar containing 3 kg/m³
MFC1, a dosage of as high as 19 weight-% SP of the cement weight, resulted in a mini-slump
flow of only 180 mm. A dosage that high is not realistic in any concrete structures, so it was
therefore decided to not increase the dosage of the SP any more. With a dosage as high as
this, the SP acts also as a retarder, and delay the cement hydration considerably. An actual
normal dosage of SP in concrete is around 1 % of the cement weight.
Comparing the plastic viscosities for this series with any of the mortars in series 1 and 2
clarifies the influence of a lower W/C ratio, which heighten the plastic viscosity.
The results are visualized in figure 5-1-13 (mini slump-flow), figure 5-1-14 (yield stress) and
in figure 5-1-15 (plastic viscosity).
Table 5-5: Test results for plasticized mortars in series 3
W/C
(-)
Plasticizer
dosage
% *
MFC
dosage
(kg/m³)
Mini
slump flow
(mm)
Yield
stress
(Pa)
Plastic
viscosity
(Pa s)
Air content
(%)
0.45 0.74 0 218 49 0.65 1.2
0.45 2.12 1.00 205 51.5 0.55 1.0
0.45 8.49 2.00 235 29.5 2.9 0.8
0.45 19.11 3.00 180 74.5 5.15 1.2
* Weight-% of cement
45
Figure 5-1-13: Mini slump flow for plasticized mortars in series 3
Figure 5-1-14: Yield stress flow for plasticized mortars in series 3
100
150
200
250
300
0,0 1,0 2,0 3,0
Min
i sl
um
p f
low
(m
m)
MFC dosage (kg/m³)
0
50
100
150
200
250
300
350
400
450
500
0,0 1,0 2,0 3,0
Yie
ld S
tres
s (P
a)
MFC dosage (kg/m³)
46
Figure 5-1-15. Plastic viscosity for plasticized mortars in series 3
5.1.4. Series 4
When comparing series 4 with series 3, it is again evident that the coarse MFC type requires
less SP than the fine MFC type. Therefore it was possible to achieve a mini slump-flow in the
target interval for all MFC dosages in series 4. Comparing the plastic viscosities in this series
with series 3, we once again can see the difference between the two MFC types. The coarse
MFC type results in a lower viscosity than the fine MFC type. Although series 3 showcase a
lower plastic viscosity with the mortar containing 1 kg/m³ than series 4, this is however
marginal and within the margin of error. All the data in series 4 is shown in table 5-6. The
results are also shown in figure 5-1-16 (mini slump-flow), figure 5-1-17 (yield stress) and in
figure 5-1-18 (plastic viscosity).
Table 5-6: Test results for plasticized mortars in series 4
W/C
(-)
Plasticizer
dosage
% *
MFC
dosage
(kg/m³)
Mini
slump flow
(mm)
Yield
stress
(Pa)
Plastic
viscosity
(Pa s)
Air content
(%)
0.45 0.74 0 218 50 0.65 1.2
0.45 1.90 1.00 227 34 0.7 1.0
0.45 5.91 2.00 245 21 1.9 0.6
0.45 17.09 3.00 227 34 4.1 0.7
* Weight-% of cement
0
0,5
1
1,5
2
2,5
3
3,5
4
4,5
5
5,5
0,0 1,0 2,0 3,0
Pla
stic
Vis
cosi
ty (
Pa s
)
MFC dosage (kg/m³)
47
Figure 5-1-16: Mini slump flow for plasticized mortars in series 4
Figure 5-1-17: Yield stress flow for plasticized mortars in series 4
100
150
200
250
300
0,0 1,0 2,0 3,0
Min
i sl
um
p f
low
(m
m)
MFC dosage (kg/m³)
0
50
100
150
200
250
300
350
400
450
500
0,0 1,0 2,0 3,0
Yie
ld S
tres
s (P
a)
MFC dosage (kg/m³)
48
Figure 5-1-18: Plastic viscosity for plasticized mortars in series 4
5.1.5. Series 5
The reference mortar in series 5 is the same recipe as the reference for series 3 and 4, but a
different mix. These two reference mortars have therefore similar test results. The dosage of
superplasticizer in the mortars with fibrils has the same dosage of superplasticizer as the
reference mortar, which target the mini-slump flow interval of 200 – 250 mm. When MFC are
added the mini-slump flow decreases as shown in table 5-7 and figure 5-1-19. It continues to
decrease with increasing MFC1 dosage.
The yield stress increases several times with an increased dosage of MFC1 as seen in figure 5-
1-20, and the plastic viscosity also increases with higher dosages of MFC1 when no
considerable amount of superplasticizer is added, figure 5-1-21.
The air content for each mortar also increases with an increasing dosage of MFC1 when no
considerable amount of superplasticizer is added, compared to series 1 to 4, which all had
about the same air content.
0
0,5
1
1,5
2
2,5
3
3,5
4
4,5
5
5,5
0,0 1,0 2,0 3,0
Pla
stic
Vis
cosi
ty (
Pa s
)
MFC dosage (kg/m³)
49
Note: Important to point out when looking at the yield stress- and plastic viscosity figures (5-
2-10 and 5-2-11 respectively) is that the scale on the y-axis is different from previous figures
in this chapter.
Table 5-7: Test results for plasticized mortars in series 5
W/C
(-)
Plasticizer
dosage
% *
MFC
dosage
(kg/m³)
Mini
slump flow
(mm)
Yield
stress
(Pa)
Plastic
viscosity
(Pa s)
Air content
(%)
0.45 0.74 0 211 62 0.9 0.9
0.45 0.74 1.00 110 238 2.5 1.4
0.45 0.74 2.00 105 440 4.7 2.5
0.45 0.74 3.00 100 738 8.3 3.8
* Weight-% of cement
Figure 5-1-19: Mini slump flow for plasticized mortars in series 5
100
150
200
250
300
0,0 1,0 2,0 3,0
Min
i sl
um
p f
low
(m
m)
MFC dosage (kg/m³)
50
Figure 5-1-20: Yield stress for plasticized mortars in series 5
(Note: The y-axis is different from previous figures)
Figure 5-1-21: Plastic viscosity for plasticized mortars in series 5
(Note: The y-axis is different from previous figures)
0
50
100
150
200
250
300
350
400
450
500
550
600
650
700
750
800
0,0 1,0 2,0 3,0
Yie
ld S
tres
s (P
a)
MFC dosage (kg/m³)
0
1
2
3
4
5
6
7
8
9
0,0 1,0 2,0 3,0
Pla
stic
Vis
cosi
ty (
Pa s
)
MFC dosage (kg/m³)
51
5.1.6. Discussion
In figure 5-1-23 it is shown how clearly the SP demand is linked to 1) the w/c, but also to 2)
the type of MFC. The lower w/c results in considerably more SP dosage needed for the same
mini slump-flow as the higher w/c. For the fine MFC type more SP is needed for the same
mini slump flow, relative to the course MFC type. Note that the SP dosage for the mortar at
w/c 0.45 with 3 kg/m³ MFC1 would have been higher for the mini-slump flow to reach higher
than 180 mm.
Beside the dubious plastic viscosity value for the mortar at w/c 0.60 with 2 kg/m³ MFC1, the
trend of increasing plastic viscosity with increasing MFC dosage is clearly visible. This is
shown in figure 5-1-24 were also the difference between MFC type and w/c is obvious. Note
that 1 kg of any MFC type does not affect the plastic viscosity to any significant extent.
Which means that lower dosages of fibrils are not needed to research, in regard to rheology.
The air volume in the mortars are varying from series to series, MFC dosage to MFC dosage,
so no real conclusion can be said about the air volume in series 1-4. It is however low in all
series 1 to 4. When proportioning concrete it is usually common to expect an air volume of
2%. In series 5 though, with a constant dosage of SP, the air volume increases with an
increasing MFC dosage. This is no surprise since the cellulose fibrils are porous in its nature
and therefore a higher air volume is to be expected in the mortars with higher MFC dosages.
The increasing viscosity, resulting from increasing MFC dosages, is interesting but rather
academic. In practice, the increasing viscosity is gained at to high cost for the necessary SP
dosage to level the flow ability to that of the mortar without MFC.
There is a strong correlation between yield stress and mini-slump flow, this is shown in figure
5-1-22 where results for all our tested mortars are plotted. The tests follow a clear trend and
therefore strengthen the validity of our rheology tests.
In figure 5-1-24 the MFC dosage is plotted against the yield stress. The MFC dosage has been
regulated after the slump flow, and therefore the yield stress show similar numbers.
In figure 5-1-25 the plastic viscosity is plotted against the yield stress for series 5. Here it is
shown that with increased plastic viscosity the yield stress also increases, as expected.
52
Figure 5-1-22: Correlation between yield stress and mini slump-flow for all tested mortars
(except series 5)
Figure 5-1-23: Correlation between fibril dosage and SP dosage, series 1-4.
(Note: Series 3 with 3 kg/m3 fibrils do not reach the slump flow target intervall)
y = 1579,6e-0,017x
R² = 0,89
0
50
100
150
200
250
300
350
100 150 200 250 300 350
Yie
ld s
tres
s (P
a)
Mini slump-flow (mm)
0,74
2,12
8,49
19,11
0
2
4
6
8
10
12
14
16
18
20
0,0 1,0 2,0 3,0
Su
per
pla
stic
izer
(%
of
cem
ent
wei
gh
t)
Fiber dosage (kg/m³)
0,45 (MFC1)
0,45 (MFC2)
0,60 (MFC1)
0,60 (MFC2)
53
Figure 5-1-24: Yield stress for all mortars in series 1- 4
Figure 5-1-25: Yield stress for all mortars in series 1-5
0
50
100
150
200
250
300
350
400
450
500
0 1 2 3
Yie
ld S
tres
s (P
a)
MFC dosage (kg/m³)
Series 1 (0.60 MFC1)
Series 2 (0.60 MFC2)
Series 3 (0.45 MFC1)
Series 4 (0.45 MFC2)
0
0,5
1
1,5
2
2,5
3
3,5
4
4,5
5
5,5
0 1 2 3
Pla
stic
Vis
cosi
ty (
Pa s
)
MFC dosage (kg/m³)
Series 1 (0.60 MFC1)
Series 2 (0.60 MFC2)
Series 3 (0.45 MFC1)
Series 4 (0.45 MFC2)
54
Figure 5-1-26: Plastic viscosity plotted against yield stress in series 5
0
100
200
300
400
500
600
700
800
0 2 4 6 8 10
Yie
ld S
tres
s [P
a]
Plastic Viscosity [Pa s]
55
5.2. Mechanical properties
5.2.1. Compressive and flexural strength
The Compressive and flexural strength tests are carried out at 1, 7 and 28 days of age. If the
mortars have not been able to be demoulded at day one (the day after moulding) because they
haven’t hardened yet, the compressive and flexural strength tests are carried out the day the
mortars are hard enough and are able to be demoulded. Only a few of the mortars have been
able to be demoulded the first day. That is often because of the high dosage of superplasticizer
needed due to the high amount of MFC in the mortars. The superplasticizer has been added to
the mortar to result in a mini-slump flow between 200 and 250 mm. Because of the high
dosage of MFC in some of the mortars, a high dosage of SP is needed to reach a mini-slump
flow in between this range. The high dosage of the SP have therefore been acting as a retarder
in those mortars, and the result of that is a much longer time for the mortar to harden.
5.2.1.1. Series 1
In figure 5-2-1 the flexural strength tests for series 1 is presented. The mortars containing
fibrils in series 1 begin at day two in the diagram. The reason for this is because they had not
yet hardened at day one, and therefore they had to be demoulded the second day. The same
mortars are tested for flexural (and compressive) strength on day eight instead of day seven.
This is because of the testing apparatus were occupied on day seven, but free to use the day
after.
Also seen is that after two days the mortars with 1 and 2 kg/m³ fibrils have the same flexural
strength, and that after eight days the mortar with 2 kg/m³ fibrils have about 14% higher
flexural strength than the mortar with 1 kg/m³ fibrils. The mortar with 3 kg/m³ fibrils exhibits
the same flexural strength after 8 days as the mortar with 1 kg/m³ fibrils after 8 days.
After 28 days the mortars containing 2 and 3 kg/m³ fibrils showcase the same flexural
strength that is higher than the reference mortars and the mortars containing 1 kg/m³ fibrils.
In figure 5-2-2 all four different mortars start out with different compressive strengths. All
four of them though have almost identical strengths after seven resp. eight days, and so is the
case after 28 days. Only the reference mortar has a slightly lower compressive strength than
the others.
The deviations for the 28 day strength readings are as follows:
Table 5-2-1: Deviations for flexural strength, series 1 after 28 days
Dosage Min Average Max
[%] [Mpa] [Mpa] [Mpa] [%]
Ref 1,73 7,38 7,51 7,64 1,73
1 kg/m³ 4,57 7,52 7,88 8,25 4,70
2 kg/m³ 8,18 8,09 8,80 9,28 5,45
3 kg/m³ 2,92 8,65 8,91 9,16 2,92
56
Table 5-2-2: Deviations for compressive strength, series 1 after 28 days
Dosage Min Average Max
[%] [Mpa] [Mpa] [Mpa] [%]
Ref 6,49 37,32 39,91 41,60 4,09
1 kg/m³ 4,29 39,71 41,49 42,65 2,82
2 kg/m³ 3,60 40,19 41,69 43,35 3,98
3 kg/m³ 1,44 41,13 41,73 42,56 1,97
Figure 5-2-1: Flexural strength for series 1, w/c 0.60 MFC1
0
1
2
3
4
5
6
7
8
9
10
11
12
0 7 14 21 28
Fle
xu
ral
stre
ngth
[M
Pa]
Time [days]
Ref
1 kg/m³
2 kg/m³
3 kg/m³
57
Figure 5-2-2: Compressive strength series 1, w/c 0.60 MFC1
5.2.1.2. Series 2
Shown in figure 5-2-3 are the flexural strength tests for series 2 (with the coarse fibril type).
This series were unlike series 1 (and 3) possible to demould after one day. The reason for that
can be explained by the much lower needed SP dosage when the coarse fibril type is used.
This is especially the case in the mortars containing 2 and 3 kg/m³ fibrils, which both have a
lower dosage of superplasticizer than the mortars with 2 kg/m³ fibrils in series 1.
The starting points on day one for the mortars in this series regarding flexural strength appears
to be of the same order as the mortars did for flexural strength in series 1.
Here however the reference mortars exhibit the highest flexural strength after seven days,
followed by the mortars containing 1 and 3 kg/m³ fibrils respectively. The curves for the
reference mortar and the mortar with 3 kg/m³ fibrils are linear, as is the case with the mortars
with 1 and 2 kg/m³ fibrils.
At the age of 28 days very little differentiates between the mortars concerning flexural
strength in series 2.
In figure 5-2-4 the compressive strength developments are shown. Very little difference can
be seen between the mortars. What can be said though is that the mortars with 3 kg/m³ fibrils
have a lower compressive strength at day one than the others. At day seven the reference
mortar appear to have the highest compressive strength, while it after twenty eight days
0
10
20
30
40
50
60
70
0 7 14 21 28
Com
pre
ssiv
e st
ren
gth
[M
Pa]
Time [days]
Ref
1 kg/m³
2 kg/m³
3 kg/m³
58
showcase the lowest compressive strength and the mortars containing 1 kg/m³ have the
highest strength. This is however marginal differences.
Since these are trials with a certain error margin and the mortars showcase almost identical
compressive strength after twenty eight days, you cannot really say one type has more
strength after 28 days than another type.
The deviations for the 28 day strength readings are as follows:
Table 5-2-3: Deviations for flexural strength, series 2 after 28 days
Dosage Min Average Max
[%] [Mpa] [Mpa] [Mpa] [%]
Ref 1,73 7,38 7,51 7,64 1,73
1 kg/m³ 8,91 7,36 8,08 8,67 7,30
2 kg/m³ 6,45 7,41 7,91 8,44 6,57
3 kg/m³ 2,72 7,52 7,73 8,06 4,27
Table 5-2-4: Deviations for compressive strength, series 2 after 28 days
Dosage Min Average Max
[%] [Mpa] [Mpa] [Mpa] [%]
Ref 6,49 37,32 39,91 41,60 4,09
1 kg/m³ 3,45 32,44 33,60 34,96 4,08
2 kg/m³ 1,88 36,51 37,21 38,45 3,33
3 kg/m³ 2,27 38,68 39,58 40,41 2,10
59
Figure 5-2-3: Flexural strength series 2, w/c 0.60 MFC2
Figure 5-2-4: Compressive strength series 2, w/c 0.60 MFC2
0
1
2
3
4
5
6
7
8
9
10
11
12
0 7 14 21 28
Fle
xu
ral
stre
ngth
[M
Pa]
Time [days]
Ref
1 kg/m³
2 kg/m³
3 kg/m³
0
10
20
30
40
50
60
70
0 7 14 21 28
Com
pre
ssiv
e st
ren
gth
[M
Pa]
Time [days]
Ref
1 kg/m³
2 kg/m³
3 kg/m³
60
5.2.1.3. Series 3
In figure 5-2-5 the flexural strength for series 3 are shown. Note that the first measurements
was done after two days of age for the mortars with 2 kg/m³ MFC1 and after three days of age
for the mortars with 3 kg/m³ MFC1. This is due to the large dosage of superplasticizer acting
as a retarder in these mortars. Therefore these mortars could not be demoulded until after two
and three days respectively.
After 7 days of age 1 kg/m³ and 2 kg/m³ show the highest value for flexural strength.
Flexural strength measured after 28 days of age showed a lower strength than after 7 days of
age for all different dosages of MFC, including the reference mortar. This is however not
supported by any available hypothesis, so the result shall not be regarded as significant.
In figure 5-2-6 the compressive strength for series 3 are shown.
Measurements made after 7 days of age show that 1 kg/m³ and 2 kg/m³ MFC1 has the highest
value for compressive strength.
After 28 days of age 1 kg/m³ and reference mortar show the highest value for compressive
strength.
Note that 1 kg/m³ had the highest compressive strength from day 1 to day 28.
However the differential margin is once again small.
The deviations for the 28 day strength readings are as follows:
Table 5-2-5: Deviations for flexural strength, series 3 after 28 days
Dosage Min Average Max
[%] [Mpa] [Mpa] [Mpa] [%]
Ref 16,25 5,98 7,14 8,58 20,17
1 kg/m³ 5,48 7,08 7,48 7,88 5,21
2 kg/m³ 4,52 6,12 6,41 6,80 6,08
3 kg/m³ 14,31 5,51 6,43 7,59 18,04
Table 5-2-6: Deviations for compressive strength, series 3 after 28 days
Dosage Min Average Max
[%] [Mpa] [Mpa] [Mpa] [%]
Ref 8,09 52,06 56,63 58,52 3,32
1 kg/m³ 8,57 52,18 57,07 59,85 4,87
2 kg/m³ 4,98 51,48 54,18 59,53 9,87
3 kg/m³ 5,48 46,53 49,23 53,92 9,53
61
Figure 5-2-5: Flexural strength series 3, w/c 0.45 MFC1
Figure 5-2-6: Compressive strength series 3, w/c 0.45 MFC1
0
1
2
3
4
5
6
7
8
9
10
11
12
0 7 14 21 28
Fle
xu
ral
stre
ngth
[M
Pa]
Time [days]
Ref
1 kg/m³
2 kg/m³
3 kg/m³
0
10
20
30
40
50
60
70
0 7 14 21 28
Com
pre
ssiv
e st
ren
gth
[M
Pa]
Time [days]
Ref
1 kg/m³
2 kg/m³
3 kg/m³
62
5.2.1.4. Series 4
In figure 5-2-7 flexural strength for series 4 are shown.
After the first measurement the mortar with 2 kg/m³ fibrils have the highest flexural strength
while the mortar with 3 kg/m³ fibrils has the lowest flexural strength.
After 28 days the mortars with 2 and 3 kg/m³ fibrils have almost identical flexural strength.
The mortars with 1 kg/m³ fibrils and the reference mortars though appear to have a lower
flexural strength than after 7 days. This is however once again not supported by any available
hypothesis, so the result shall not be regarded as significant.
In order to receive reasonable results we deleted results from one of the three mortars
containing 1 kg/m³ fibrils at day 7. This was because of the reading of that specific mortar
where half as much of any of the other two mortars tested, and much less than the readings for
the 7 day strength tests.
The same have been done with two of the three mortars containing 2 kg/m³ fibrils at day 7.
At day 28 two of the three mortars containing 3 kg/m³ were deleted because of dubious
readings. The readings of the specimens were not as low as half of the strength of the third
mortar, but low enough that the mean reading of these mortars would be lower than the 7 day
readings, and it was decided that they should be deleted.
In figure 5-2-8 compressive strength tests results for series 4 are shown.
Values for the 1 and 7 days of age concerning compressive strength are all about the same for
all samples except for the mortars containing 3 kg/m³ fibrils, which can be explained by the
large dosage of the superplasticizer.
All of the mortars though are in the range of the measurement error margin.
At the age of 28 days all of the mortars containing fibrils appear to have the same
compressive strength. The reference mortar has a lower compressive strength than the mortars
containing fibrils.
Compressive strength tests were done after 2 days of age for both the reference mortars and
the mortars containing 3 kg/m³ fibrils. This was due to the large dosage of superplasticizer in
the mortar containing 3 kg/m³ which cause an extension to the time it takes for the hardening,
and because the testing apparatus on day seven for the reference mortar were occupied.
The deviations for the 28 day strength readings are as follows:
Table 5-2-7: Deviations for flexural strength, series 4 after 28 days
Dosage Min Average Max
[%] [Mpa] [Mpa] [Mpa] [%]
Ref 16,25 5,98 7,14 8,58 20,17
1 kg/m³ 1,03 8,60 8,70 8,79 1,03
2 kg/m³ 3,51 9,61 9,96 10,36 4,02
3 kg/m³ 0 9,73 9,73 9,73 0
63
Table 5-2-8: Deviations for compressive strength, series 4 after 28 days
Dosage Min Average Max
[%] [Mpa] [Mpa] [Mpa] [%]
Ref 8,09 52,06 56,63 58,52 3,32
1 kg/m³ 1,48 61,79 62,72 64,26 2,46
2 kg/m³ 3,50 61,69 63,85 68,42 7,16
3 kg/m³ 0,16 62,47 62,57 62,66 0,14
Figure 5-2-7: Flexural strength series 4, w/c 0.45 MFC2
0
1
2
3
4
5
6
7
8
9
10
11
12
0 7 14 21 28
Fle
xu
ral
stre
ngth
[M
Pa]
Time [days]
Ref
1 kg/m³
2 kg/m³
3 kg/m³
64
Figure 5-2-8: Compressive strength series 4, w/c 0.45 MFC2
5.2.1.5. Series 5
In figure 5-2-9 flexural strength test results for series 5 are shown. Tests were only performed
at 28 days of age in series 5, and therefore the figure only shows the flexural strength at 28
days of age. The results show a reduction in flexural strength with increased amount of fibrils
in the mortars, the reference mortars has the highest flexural strength while the mortars with 3
kg/m³ fibrils has the lowest flexural strength.
In figure 5-2-10 compressive strength for series 5 are shown. The figure shows the
compressive strength at 28 days of age. The results for compressive strength follow the same
pattern as the flexural strength tests, an increased amount of fibrils lead to a reduction in
compressive strength. The difference is however not big, except for the mortar with 3 kg/m3
fibrils.
The deviations for the 28 day strength readings are as follows:
Table 5-2-9: Deviations for compressive strength, series 5 after 28 days
Dosage Min Average Max
[%] [Mpa] [Mpa] [Mpa] [%]
Ref 7,06 51,37 55,27 60,00 8,56
1 kg/m³ 8,87 49,19 53,98 59,07 9,43
2 kg/m³ 3,78 51,11 53,13 55,78 5,01
3 kg/m³ 0,63 45,71 46,00 46,22 0,48
0
10
20
30
40
50
60
70
0 7 14 21 28
Com
pre
ssiv
e st
ren
gth
[M
Pa]
Time [days]
Ref
1 kg/m³
2 kg/m³
3 kg/m³
65
Table 5-2-10: Deviations for flexural strength, series 5 after 28 days
Dosage Min Average Max
[%] [Mpa] [Mpa] [Mpa] [%]
Ref 3,86 8,23 8,55 8,91 4,09
1 kg/m³ 13,14 7,34 8,45 9,26 9,59
2 kg/m³ 2,07 8,06 8,23 8,55 3,89
3 kg/m³ 11,56 6,96 7,87 8,60 9,28
Figure 5-2-9: Flexural strength for series 5, w/c 0.45 MFC1, at the age of 28 days
0
1
2
3
4
5
6
7
8
9
10
11
12
Fle
xu
ral
stre
ngth
[M
Pa]
Ref
1 kg/m³
2 kg/m³
3 kg/m³
66
Figure 5-2-10: Compressive strength for series 5, w/c 0.45 MFC1, at the age of 28 days
5.2.2. Discussion
Our test results show that a w/c of 0.45 compared to 0.60 results in a higher overall strength.
In other words a lower w/c results in higher strength of the mortars than a higher w/c does.
While this is commonly known, the results in difference in strength with the addition of
cellulose fibrils in mortars are not.
The lower strength in the first strength tests for series 3 and 4 after 1-3 days for the higher
fibril dosages can be explained by the large dosage of superplasticizer acting as a retarder in
these cases. After 7 days the mortars with a high dosage of fibrils in series 3 and 4 have still
not caught up with the mortars containing lower fibril dosage or the reference mortars. This
can still be explained by the large dosage of the superplasticizer. The dosage of
superplasticizer in series 3 and 4 is higher than the dosage in series 2 and considerably higher
than in series 1, this is seen in figure 5-1-23.
The strength after 28 days in series 1 and 2 are almost the same for both flexural strength and
compressive strength, except for the flexural strength in series 1. Our results in flexural
strength for the 2 and 3 kg/m³ fibril mortars are approximately 10-15 % higher than for the
reference mortar and the mortar containing 1 kg/m³ fibrils respectively.
Otherwise the fibrils don’t seem to make any differences in the mortars in series 1 and 2.
In series 3 and 4 the results after 28 days is somewhat dubious. In series 3 both flexural
strength and compressive strength test results show that no matter the dosage of fibrils the
same amount of strength is measured. But when the compressive strength increases the
0
10
20
30
40
50
60
70C
om
pre
ssiv
e st
ren
gth
[M
Pa]
Ref
1 kg/m³
2 kg/m³
3 kg/m³
67
flexural strength decreases after 28 days. These results is unfortunately hard to interpret,
something doesn’t add up.
In series 4, however, only the reference mortar and the mortar with 1 kg/m3 fibrils show a
reduction in flexural strength after 28 days, while the compressive strength increases. These
results are again hard to interpret because they seem illogical. Otherwise the strength
increases after 28 days of age compared to 7 days of age.
One thing is worth pointing out between the different fibril types. With w/c 0.60 (series 1 and
2) the fine fibril type, MFC1, show higher strength results than the courser fibril type, MFC2.
With the w/c 0.45 (series 3 and 4) however the results are the other way around. Here it is
instead the course fibril type, MFC2, that show higher strength results than the fine fibril type,
MFC1.
Results from series 5, where the intention was to rule out any other factor than the fibrils to
affect the strength of the mortars, show that an increase of fibril dosage decreases both the
flexural and compressive strength, even though it is marginal. While in series 1 to 4 the
results points to no direct effect of the fibrils, no matter the dosage, in series 5 the results
shows a somewhat clear connection between higher fibril content with lower overall strength.
The graphs show that tendency, but this is however not statistical safe to say, when the results
in series 1 to 4 are of varying kind.
68
5.3. Shrinkage
5.3.1. Free Shrinkage
All measurements were made according to the SIS standard SS-13-72-15 [27]. Two
deviations were made from standard. The first is that the specimens were moved from a room
with 100 % humidity to a room with 50 % humidity one day after moulding. The second is
that shrinkage measurements were also taken at 0, 1 and 28 days of age after demoulding.
Instead of starting to count from the day the mortars were moulded, all measurements started
to count from the day the mortars were demoulded. Therefore because of the varying time for
the mortars to harden the mortars are not tested for free shrinkage at the same age as each
other
5.3.1.1. Series 1
In figure 5-4-1 free shrinkage for series 1 are shown. Initially the most shrinkage happens in
the reference mortars, which is almost on par with the mortar containing 3 kg/m³ fibrils. After
33 days all of the mortars have shrunk almost the same amount. After 62 days and 112 days
the mortars are almost identical though.
Figure 5-3-1: Shrinkage series 1, w/c 0.60 MFC1
0
0,5
1
1,5
2
2,5
3
0 10 20 30 40 50 60 70 80 90 100 110
Sh
rin
kage
[‰]
Age [days]
Ref
1 kg/m³
2 kg/m³
3 kg/m³
69
5.3.1.2. Series 2
In series 2, figure 5-4-2, it can be seen that after 33 days all the mortars including the
reference mortars have shrunken almost the same amount. After 112 days though the
shrinkage is almost the same in all mortars.
Figure 5-3-2: Shrinkage series 2, w/c 0.60 MFC2
0
0,5
1
1,5
2
2,5
3
0 10 20 30 40 50 60 70 80 90 100 110
Sh
rin
kage
[‰]
Age [days]
Ref
1 kg/m³
2 kg/m³
3 kg/m³
70
5.3.1.3. Series 3
In figure 5-4-3 free shrinkage for series 3 are shown. All the mortars has shrunken different
amounts, in which the mortars with 2 kg/m³ fibrils has shrunken the most, and the reference
mortars has shrunken the least. Although it looks like there is a difference in shrinkage
between the mortars, it is in reality pretty small. So it can be said that they shrink the same
percentage.
Figure 5-3-3: Shrinkage series 3, w/c 0.45 MFC1
0
0,5
1
1,5
2
2,5
3
0 10 20 30 40 50 60 70 80 90 100 110
Sh
rin
kag
e [‰
]
Age [days]
Ref
1 kg/m³
2 kg/m³
3 kg/m³
71
5.3.1.4. Series 4
In figure 5-4-4 free shrinkage for series 4 are shown. The shrinkage of all mortars is almost
the same. The amount of BAC between those mortars which have the most shrinkage
compared with least shrinkage can be read to approximately 0,3 ‰-units during the 28 day
reading. As stated before, in reality this difference is small.
Figure 5-3-4: Shrinkage series 4, w/c 0.45 MFC2
0
0,5
1
1,5
2
2,5
3
0 10 20 30 40 50 60 70 80 90 100 110
Sh
rin
kage
[‰]
Age [days]
Ref
1 kg/m³
2 kg/m³
3 kg/m³
72
5.3.2. Constrained shrinkage
In table 5-4-1 results from constrained shrinkage can be seen. The shrinkage rings has
unfortunately for most of the mortars cracked during the weekend after moulding, therefore
no exact day of cracking can be established for these specimens. It looks though as an
increase of fibril dosage postpones the cracking time. This can be seen in series 1 where the
exact age of cracking for the dosage of 3 kg/m³ is on day 7, while for the reference mortar and
the mortars with the lower dosage cracks in a span of day 3 to 6. In series 3 this is also
evident. Day of cracking are for the reference mortar along with the two lower fibril dosages
all on day 3, while the day of cracking for the mortar with 3 kg/m³ fibrils is in the span of day
5 to day 9.
On the other hand it has been observed that an increase of the superplasticizer dosage as is the
case in the mortars with 3 kg/m³ fibril dosages delay the hardening of the mortars and
therefore an increased age of cracking can therefore also be explained.
However, one or two days of postponed cracking has no real significance when larger
concrete structures are casted. If on the other hand the fibrils would delay the cracking age by
say 10 days, then it could have some significance.
Table 5-3-1: Constrained shrinkage, day of cracking.
Mix/Series W/C Fibril dosage Fibril type Age of cracking
(-) (kg/m³) (-) (days)
Ref 0.60 0.60 0 N/A 3 - 6
Series 1
0.60 1 MFC 1 3 - 6
0.60 2 MFC 1 3 - 6
0.60 3 MFC 1 7
Series 2
0.60 1 MFC 2 1 - 4
0.60 2 MFC 2 2 - 5
0.60 3 MFC 2 2 - 5
Ref 0.45 0.45 0 N/A 3
Series 3
0.45 1 MFC 1 3
0.45 2 MFC 1 3
0.45 3 MFC 1 5 - 9
Series 4
0.45 1 MFC 2 3 - 7
0.45 2 MFC 2 3 - 7
0.45 3 MFC 2 4 - 8
73
5.3.3. Surface coverage
In picture 5-4-5 results from the test with using fibrils as a surface coverage are shown. The
use of fibrils as a surface coverage of the mortar resulted in the fibrils not sticking to the
surface of the mortar. The attempt to compare the results with the mortar without fibril
suspension and the mortar with plastic coverage can therefore not be performed. A way to
circumvent this could possibly be to apply the fibril suspension while the mortars still are wet
but hard enough to apply the fibrils, to possibly make it more adhesive. This is however not
realistic on concrete, because the time to apply the fibril suspension will possibly have to be at
a time when the concrete is hardened enough to be able to walk on it.
Figure 5-3-5: The look of the fibril coated slab after 4 days
5.3.4. Discussion
No test result shows that the fibrils affect the shrinkage of the mortars by any large degree. In
series 1 the mortars containing 1 and 2 kg/m3
fibrils shrinks less than the reference mortars
and the mortars containing 3 kg/m3
up until they have shrunk almost the same after about 35
days. In series 2 the mortars shrinks the same amount except for the mortars containing 3
kg/m3 fibrils which shrinks more initially, but catches up with the other mortars at the age of
about 35 days. In series 3 however the reference mortar and the mortar containing 3 kg/m3
have shrunken the least during the 35 days of measurements.
In series 4 all of the mortars are on par with each other as they all shrink the same amount.
There is no direct pattern in shrinkage between the different series, and they are all within the
measurement error margin.
After a longer test period, 112 days, the results are identical for series 1 and 2, while in series
3 and 4 the difference cannot be said to have such a big relevance.
Apparently, the fibrils don’t do any considerable change in shrinkage to the mortars either.
Since all series points toward the same strength for all dosages of fibrils together with the
reference in each series after 90 or 112 days of age, theoretically the strength development
will continue in the same direction. Therefore our trials for free shrinkage is terminated earlier
than said in associated standard.
Due to the cracking during weekends no definitive parallel can be drawn between free
shrinkage and the day the respective mortar cracks.
74
A larger dosage of fibrils in the mortars though is hinting towards a postponing age of
cracking. But this can again be the result of the superplasticizer acting as a retarder. Also
mentioned earlier is that one or two days extra days before the mortars crack has no real
significance. If the difference was 10 or 20 days then the fibril dosages would be of a bigger
relevance.
Regarding the fibril suspension used as a coating of the concrete, it seems that the fibrils
cannot be applied on hardened concrete because of its non adhesive properties. However
applying the fibril suspension when the concrete is still wet could be interesting in that the
fibril suspension might have it easier to stick to the wet mortar. This is however not realistic,
in the sense that in reality the concrete would most likely have to be hardened enough to be
able to be walked upon, for the fibril suspension to be applied.
75
5.4. Weight loss
5.4.1. Series 1
Concerning the weight loss it is apparent in figure 5-5-1 that the mortars containing 3 kg/m³
MFC1 in series one loses more weight than the other mortars containing 1 and 2 kg/m³ MFC1
and the reference mortars. This happens initially, in the first 2 – 5 days though.
Figure 5-4-1: Weight loss series 1, w/c 0.60 MFC1
0
25
50
75
100
125
0 10 20 30 40 50 60 70 80 90 100 110
Wei
gh
t lo
ss [
‰]
Age [days]
Ref
1 kg/m³
2 kg/m³
3 kg/m³
76
5.4.2. Series 2
In figure 5-5-2 series two is presented. Here the same pattern can be seen as series one except
for the mortars containing 1 kg/m³ MFC2 are the ones losing the most weight, with the
mortars containing 3 kg/3 tightly followed. The most development of difference in weight loss
though happens in the first ten days.
Figure 5-4-2: Weight loss series 2, w/c 0.60 MFC2
0
25
50
75
100
125
0 10 20 30 40 50 60 70 80 90 100 110
Wei
gh
t lo
ss [
‰]
Age [days]
Ref
1 kg/m³
2 kg/m³
3 kg/m³
77
5.4.3. Series 3
In figure 5-5-3 series three is presented. All of the mortars shrink almost the same amount.
The one that differs a little from the rest in this series though is the mortars with 2 kg/m³
fibrils. However the deviation is within the measurement error margin.
Figure 5-4-3: Weight loss series 3, w/c 0.45 MFC1
0
25
50
75
100
125
0 10 20 30 40 50 60 70 80 90
Wei
gh
t lo
ss [
‰]
Age [days]
Ref
1 kg/m³
2 kg/m³
3 kg/m³
78
5.4.4. Series 4
In figure 5-5-5 series 4 is presented. The same pattern can be seen in series 4 as in series 3
except for the mortars with 3 kg/m³ fibrils is on par with the mortars containing 2 kg/m³
fibrils.
Figure 5-4-4: Weight loss series 4, w/c 0.45 MFC2
0
25
50
75
100
125
0 20 40 60 80
Wei
gh
t lo
ss [
‰]
Age [days]
Ref
1 kg/m³
2 kg/m³
3 kg/m³
79
5.4.5. Discussion
No real patterns can be seen between different series or different MFC dosages. All of the
mortars containing MFC loses either more weight or as much weight as the reference mortars
without any MFC.
Only in series 1 does the mortars containing 3 kg/m3 out compete the other mortars in weight
loss.
Series 3 is the only series where the mortars with the highest dosage of MFC lose the least
weight, but the difference from the reference mortars and the mortars with 1 kg/m³ MFC in
this series is almost nonexistent. It is the mortars with 2 kg/m3
MFC that lose the most weight.
In series 4 the difference in weight loss is not as apparent as in series 1 and 2. It can still be
seen that there is a large enough gap to say there is a difference in weight loss with and
without MFC. In series 4 the mortars containing 2 and 3 kg/m3 MFC have a slightly higher
weight loss than the reference mortars and the mortars containing 1 kg/m3 MFC.
Also note that all four series has shrunk the same amount but series 1 and 2 has lost more
weight than series 3 and 4. This can be explained by that the drying shrinkage is bigger with
w/c of 0.60 and the autogenous shrinkage is bigger with w/c of 0.45.
With that said there are no conclusive parallels drawn between the series containing the fine
fibril fraction, MFC1, in series 1 and 3, since there is no pattern between the different fibril
dosages in these two series. There is no apparent parallels between the two series containing
the course fibril type either, MFC2, in series 2 and 4.
The biggest difference in weight loss between the MFC and the reference mortar is in series 1
and 2. This difference might seem of significance at first sight, but the difference is only
about 3-4 percentage, which in our mortars weighing half a kilo is only a few grams.
At first sight it might seem that the dosages of MFC are of importance when looking at each
series by itself. But when looking at the whole picture no real pattern evolves. The differences
in weight loss in series 1 and 2 are of no big importance, so therefore the conclusion of our
trials is that the fibrils have no or very little affect on the weight loss of the mortars.
80
5.5. Anti washout resistance
All of the mortars in the anti-washout tests have a w/c of 0.45, and the mortars containing
fibrils are of the fine fraction type, MFC1. The results of the anti washout test is presented in
table 5-6-1 and in figure 5-6-1. When casting underwater the concrete in its consistency has to
be very flowable, since compaction of the fresh concrete under water is almost impossible.
Therefore, all the tested mortars where plasticized to the same degree of flowability.
Table 5-5-1. Rheology tests results and weight loss for all of the mixtures in the anti-washout tests
Mortar
type
AWA
dosage
(kg/m³)
Plasticizer
(% )*
MFC1
dosage
(kg/m³)
Mini
slump
flow
(mm)
Yield
stress
(Pa)
Plastic
viscosity
(Pa s)
Air
Content
(%)
Mass loss
after 9
lifts
(%)
AW 2,00 1.32 0 224 32 2.2 0.9 3.40
REF 0 0.74 0 200 61 0.7 1.1 7.03
1 kg/m³ 0 2.12 1,00 203 49 1.4 0.8 8.19
2 kg/m³ 0 8.49 2,00 228 35 2.6 0.4 9.32
3 kg/m³ 0 19.11 3,00 180 70 5.1 0.7 8.55
* Weight-% of cement
Figure 5-5-1: Results from the anti-washout tests
0%
1%
2%
3%
4%
5%
6%
7%
8%
9%
10%
0 1 2 3 4 5 6 7 8 9
Per
cen
tage
of
wash
ed o
ut
mort
ar,
of
init
ial
wei
gh
t
Number of Immersions
AWA
REF
1kg
2kg
3kg
81
5.5.1. Discussion
The results are clear in that the fibril suspensions are not working as an anti-washout
admixture. From the first immersion the mortars containing fibrils loses way more of its
contents than the mortar containing a regular AWA. The mortars containing fibrils is even
worse in keeping the contents of the mortars than the reference mortar with no addition of
fibrils.
5.6. Capillary porosity
The trials were performed according to the standard SS-EN ISO 15148 [21], with a few
deviations as mentioned in section 3.3.5.
When the line in the graph shows a sudden decrease in slope it means that water has appeared
on the top of the surface of the specimen, and the water absorption coefficient can be
calculated with a value of the time, tf, less than a day. The 24 h readings are therefore not used
in calculating the water capillary coefficient. To make the calculations easier and because of
the fact that no readings have been done after 8 h, the time, tf, is set to 8 h for all specimens.
The Δmt plotted against √ for series 1 is shown in figure 5-1, (only series 1 will be shown as
an example).
Figure 5-1. Example of plotted results of Δmt = (mt – mi)/A, against √ for series 1, w/c 0.60 MFC1
A line is drawn from the time tf, which is 8 h, through the plotted graph and extend it to zero
time, Δmʹ0, as shown in figure 5-2.
0
1
2
3
4
5
6
7
8
0 50 100 150 200 250 300
Dif
fere
nce
bet
wee
n l
ost
mass
of
each
wei
gh
ing a
nd
div
ided
wit
h t
he
are
a
(Δ
mt
= kg
/m²)
Square root of time ( )
Ref
1 kg/m³
2 kg/m³
3 kg/m³
82
Figure 5-2. Example of the same graph as in figure 5-1, but here with trend lines and the 24 h
measurements not included.
As mentioned in section 3.3.5 the water capillary coefficient is calculated by
Aw = (Δmʹtf - Δmʹ0) / √ f
where
Δmʹtf is the value of Δm on the straight line at time tf, in kg/m2
tf is the duration of the test, in this case at 8 h, in seconds
In our trials, however, we have read the water capillary coefficient from the slope of the line,
which is calculated by excel with a second grade equation, seen in figure 5-2.
The conclusion from the water capillary porosity tests is that the higher the dosage of fibrils in
the mortars, the more dense the mortars get, and the higher the resistance is to water
absorption. The results are shown in figure 5-3. The reason for a lower water coefficient with
higher dosages of micro fibrilar suspension could be that higher dosages of fibrils densifies
the structure. Series 1,3 and 4 showcase similar graphs. Series 2 however deviates from the
results of the other series. If something went wrong with the procedures is hard to tell, or if
the samples were somehow defect, but considering the similarity of the other series something
is not right with the 2nd
series. More samples than one per type would be preferable, however
this was not the case here, we only had one sample per type. This was due to lack of time and
moulded mortar.
y = 0,0385x + 0,1575
R² = 0,9977
y = 0,0322x + 0,099
R² = 0,998
y = 0,0235x + 0,1274
R² = 0,997
y = 0,0195x + 0,1321
R² = 0,9966
0
1
2
3
4
5
6
7
8
0 50 100 150 200 250 300
Dif
fere
nce
bet
wee
n l
ost
mass
of
each
wei
gh
ing a
nd
div
ided
wit
h t
he
are
a
(Δm
t =
kg
/m²)
Square root of time ( )
Ref
1 kg/m³
2 kg/m³
3 kg/m³
83
Figure 5-3. Water absorption coefficient for all four series
When looking at the actual specimens, figure 5-4, after the tests have been conducted for 24 h,
it can clearly be seen on the specimens with higher micro fibrilar dosages that almost no damp
spots have appeared on the upper surfaces.
Figure 5-4. Specimens after 24 h in the water tank
0
50
100
150
200
250
300
350
400
450
0 1 2 3
Wa
ter
ab
sorp
tio
n c
oef
fici
ent
10
¯⁴
(kg/(
m²sˉ¹
))
Fiber dosage (kg/m³)
Series 1 (w/c 0.60 MFC1)
Series 2 (w/c 0.60 MFC2)
Series 3 (w/c 0.45 MFC1)
Series 4 (w/c 0.45 MFC2)
84
85
6. Conclusions
The conclusions from the tests performed in this thesis can generally be summarized as the
evaluated cellulose fibrils have no significant effect as mortar reinforcement. As an rheology
enhancer in low dosages it doesn’t have much of an effect either.
Regarding the rheology of the mortars, a higher dosage of fibrils, increases the need for a
higher dosage of the superplasticizer. The need for SP in the mortars containing 3 kg/m³
fibrils is so high that it is highly uneconomical using those dosages. If the dosage of SP isn’t
high enough in the mortar, it will be relatively stiff. Normal quantities of SP in concrete is
about 1 % of the cement weight, which most of our mortars significantly surpasses.
The results from the flexural and compressive strength tests in series 1 and 2 show no
significant effect of the fibrils in the mortars. The test results are so similar between the
reference mortars and the mortars with fibrils in series 1 and 2 that it is not possible to say that
one fibril dosage is better than another, or having fibrils in the mortar at all.
Regarding the fibril type though, the fine fraction type of the fibrils show a little increase in
strength compared to the course fibril type, both for flexural strength and compressive
strength in series 1 (fibril type MFC1) and in series 2 (fibril type MFC2)
Series 3 and 4 (w/c 0.45) show somewhat dubious results in the sense that the flexural
strength decreases in both series 3 and 4, except for the 2 and 3 kg/m3 fibril dosages in series
4. Concerning the strength between the two different fibril types the readings are vice versa
from series 1 and 2 (w/c 0.60). Here instead, the course fibril type, MFC2, show higher
strength results than the finer fibril type, MFC1.
The cellulose fibrils show no effect on the free shrinkage. However, it seems that with the w/c
of 0.60 (series 1 and 2), the specimens shrink the same amount after 35 days and further, but
the mortars containing fibrils initially shrinks less than the reference mortars. The mortars
with a w/c of 0.45 seem to neither shrink more than the reference mortars initially nor shrink
differently totally, from the reference mortars. The mortars vary a little in shrinkage in series
3 and 4, but this is not significant.
Concerning the weight loss test, the conclusion from these measurements is that addition of
fibrils in mortar either increase the weight loss of the mortars or stay the same as the reference
mortars.
The results from the constrained shrinkage tests shows that the rings cracked around the same
time and therefore shows that the fibrils do not have any significant effect on constrained
shrinkage.
The experiment with the fibril suspension acting as a surface coverage showed that the fibril
suspension could not be used as a covering film the way it was done here since the fibril
suspension didn’t stick to the surface of the mortar slabs.
When testing the fibril suspension as an anti-washout admixture it became evident that using
the fibrils for that purpose did not show any positive effects.
The results from the water absorption tests though showed an increasing resistance to
absorption with an increasing dosage of fibrils.
86
With the results from our tests with the dosages we’ve used, it can be concluded that use of
these fibril suspensions is not recommended other than for the purpose of making the concrete
more resistant to water absorption.
87
7. References
[1] Wedin, I. (2011). Introduction meeting with project manager Irene Wedin from Stora
Enso. 12 April 2011 (approximately 2h)
[2] R.F. Zollo. (1997). Fiber-Reinforced Concrete: An Overview After 30 Years of
Development. Cement and Concrete Composites. Vol. 19. Iss. 2. pp. 107-122
[3] Y.-M. Chun and T.R. Naik. (2004). Repulping Fibrous Residuals from Pulp and Paper
Mills for Recycling in Concrete. Tappi Journal. Vol. 3. iss. 12. pp. 7-10.
[4] R. MacVicar, L.M. Matuana and J.J. Balatinecz. (1999). Aging Mechanisms in Cellulose
Fiber Reinforced Cement Composites. Cement and Concrete Composites. Vol.21. iss. 3. pp.
189-196.
[5] M.A.S. Mohamed, E. Ghorbel and G. Wardeh. (2010). Valorization of micro-cellulose
fibers in self-compacting concrete. Construction and Building Materials, vol. 24, No. 12, pp.
2473-2480
[6] Zhengwu Jiang, Nemkumar Banthia and Sarah Delbar. (2009). Effect of Cellulose Fiber
on Properties of Self-compacting Concrete with High-volume Mineral Admixtures. Second
International Symposium on Design, Performance and Use of Self-Consolidating Concrete,
SCC’2009-China, June 5-7 2009, Beijing, China
[7] SS-EN 196–1:2005 Swedish Standard Institute (2005). Method of testing cement - Part 1:
Determination of strength. 2nd
Ed. pp. 33.
[8] SS-EN 932-2. Swedish Standard Institute (2007). Del 2: Neddelning av laboratorieprov.
1st Ed.
[9] Information about the operation of CBI. Taken from
http://www.cbi.se/viewNavMenu.do?menuID=120 on the 4th
of April 2011.
[10] Bjurström, Per Gunnar. (2008). Byggnadsmaterial, Uppbyggnad, Tillverkning och
Egenskaper. E.d.2:3. Pozkal, Poland: Studentlitteratur. 562 pp. ISBN 978-91-44-02738-8.
[11] Ljugkrantz C, Möller G, Petersons N (1994). Betonghandbok Material. Ed. 2. pp 1127.
Stockholm. AB Svensk Byggtjänst. ISBN 91-7332-709-3.
[12] Neville A.M. (1995). Properties of Concrete. 4th
Ed. pp. 844. Edinburgh. Harlow. Essex.
Pearson Education Limited. ISBN: 0-582-23070-5
[13] Wallevik O.H. (2009). Introduction to Rheology of Fresh Concrete. Reykjavik, Iceland.
Chapter 5-6.
[14] Ramachandran V.S (1995). Concrete Admixtures Handbook, Properties, Science, and
Technology. 2nd
Ed. pp 1153. Ottawa, Ontario, Canada. National Research Council Canada.
ISBN: 0-8155-1373-9
88
[15] Westerholm M (2006). Rheology of the Mortar Phase of Concrete with Crushed
Aggregate. Luleå. Luleå University of Technology. pp 94.
[16] Billberg P. (1999). Some rheology aspects on fine mortar part of concrete. Lic. Thesis,
Royal institute of Technology Department of Structural Engineering. TRITA-BKN. Bulletin
51. ISRN KTH/BKN/B—51---SE
17[16] M. Al-Emrani, B. Engström, M. Johansson and P. Johansson. (2008). Bärande
Konstruktioner Part 1. publ. 2008:12. Chalmers Tekniska Högskola. Göteborg. ISSN 1652-
9162
18[18] F.C. Jorge, C. Pereira and J.M.F. Ferreira. (2004). Wood-Cement Composites: A
Review. International Journal of Advanced Manufacturing Technology. Vol. 24. Iss. 7-8. pp.
370-377.
19[19] Carsten Vogt, Hans Hedlund, Kjell Wallin, Franziska Baldy and David Petersson.
(2010). Beständiga undervattensgjutna kajkonstruktioner. SBUF projekt 11940. pp. 162.
20[19] M.Sonebi and K.H. Khayat. (2001). Effect of Mixture Composition on Relative
Strength of Highly Flowable Underwater Concrete. ACI Materials Journal, vol. 98, No. 3, pp.
233-239
21[21] Kamal Henri Khayat. (1995). Effects of Antiwashout Admixtures on Fresh Concrete
Properties. ACI Materials Journal, vol. 92, No. 2, pp. 164-171
22[22] SS-EN ISO 15148. Swedish Standard Institute (2011). Fukt- och värmetekniska
egenskaper hos byggmaterial och byggprodukter – Bestämning av kapillaritetskoefficient
genom partiell nedsänkning i vatten. 1st Ed.
23[23] Cementa AB (2011). Byggcement CEM II/A-LL 42,5 R. Technical data sheet
downloaded on the 4th
of April 2011.
24[24] SS-EN-197-1 Swedish Standard Institute. (2000). Cement – Del 1: Sammansättning
och fordringar för ordinära cement. 1st
ed.
25[25] T.K. Erdem, K.H. Khayat and A. Yahia. (2009). Correlating Rheology of SCC to that
of Corresponding Concrete-Equivalent Mortar. ACI Materials Journal. vol. 106. no. 2. pp.
154-160
26[26] SS-13-72-15. Swedish Standard Institute (1978). Betongprovning – Hårdnad Betong -
Krympning. Ed. 1
27[27] ASTM Standard. C 1581-04. Standard Test Method for Determining Age at Cracking
and Induced Tensile Stress Characteristics of Mortar and Concrete under Restrained
Shrinkage.
28[28] ASTM Standard. CRD-C 61-89A. Test method for determining the resistance of
freshly mixed concrete to washing out in water. Issued 1 Dec 1989.
89
Appendix o Byggcement CEM II/A-LL 42.5 R, Product sheet
o Sikament ECO 12-2, Product sheet
o Sika UCS, Product sheet
��������������������� ������������������������������������������� ��!����� �"�#!�����$�"� ��%�&���$������ ����$'��(���$���)�!���*� �����$���!�����! ��!��� ���+�������+����,��%�-%�! �$)�+������%�.��� ���������'���/�����'�� ����0� ����������)��+���%��1��1�������1��������������)��(���������!������%�""*23�456*47�����*&���48�"���������� �����$�(���� �����(#���� �����������%������������"�#!������9*�������������� /���*�)�����4:4;4<���$�72*�������������� /����)�*����4:4;0<%�����������"� ������9*�������������� /���*�)�����4:4;4=���$�72*�������������� /����)�*����4:4;0;%>?1����@�AB�C@ �1@DE�@��������A1@�AF�!����������� ��#��! ���G+)��H,�4:::����������������������0=*��������! ��I�- ��J�%�F�!��������'���(����'��!�!'���� �� ���%F��� ��� �� �K#������������������� ����!�()�� ��)(��������� ���������������������()���������!���������*�����������������������++��G�����,�!�����H% �����������('��������� ��������$������ ��� �$#����L��'�������(������!�������(�����������)�(#����������*���)�� ������)����� !��(#�������� ���((���%�"������������('��(#�!������'������������� �������� �$#����L��'�������(��������� ��������%�&���'� ��������� ��#����������� �������$��)����! ����������� ��������������������$������������������ ���� ���� ������%����M������N�1�� D� �M���� �1O#����!�������!�������������������J�I�50���$7��������PQRSTUVWXYZUVVT[R% M�1���N@C�����1@�AJ����������������(#�!������'����� ���(#��+������$���(��� �������(#�����%�7������ �#����������� ���(#�����!��*� ���#���������%�O)�� ����������+ ��������� )�$����*- ������!������ �� ���������'�$)����$����� ��������%O#��()������� �� �(����� �����$���������! �� ��������PQT\R]\SWXX_VXW%�1�CE�@�>�1���@��O#����������)������� ����!�R[WaT_VXW,�PQT\R]\SWXXb_VXW��$�cVVdQeeXYV\Z\RXeS_\SQdd\VS\Rf����!'��$��*� ���ggg%�������%��%hijjklmlnopqrsttuvwxxsyz{|s} ~)� �0::6
������������� ��������� ��������������������� �����������!���� "�"����#�$%��&��'������!(%�)��%&*!�+++(������!(%�,-�&���� ��.�/0.��1���2344567689:;<=:>>?@ABB:CDEF:G�HII���������JK*�������&���-$$�H���L���*!�%�M��*��*!�!(�N&*���%�!�!��&���O������!��KOL*��!!KK&�!��LM��!�I&K�!�%$!�������L�ILO�%�L(���PQRSTUV� WXTYZ[\]Q� WXTYZ[\]Q� 0VURR 0\Y .\UZ _[UR]Q�0TaZ]Q 0XYQ 0YUR]U\]�$��&b��H�! �"� ��� c"� ��d�I�&�*��&* ��� ��� c"� �&� ef�� #g��h ���LH���M���!%���� �*i �" �� c" jk!�LH���M���!%���� �*i " "" c" jk! ef���� #g��h ���LH���M���!%���� �*i �� c� jk!�LH���M���!%�������*i � c� jk! ef��� ����lef��� #g��h ��m�$!��*��%&��� "��� "��� c�� �Id�"��LH�*��%&��� �� � �� � c� � �Id�"no-%��� �� "� c� p�PQRSTUV� _\[RSZ[\]QR�_\[RSZ[\]QR� 0\Y .\UZ _[UR]Q�0TaZ]Q 0XYQ 0YUR]U\]j!I��%&-�&*��jIq ���r��" ��"r�� p sf ���p #g��h ���-��!��!���%���q" "��r��� "��r"� p sf����p #g��h ��m�L&*��t�uu ����r���� ���"r��� p sf�����p #g��h ��N!�����J%�&I�tL�v �r� �r� $$� sf��$$��w���(�#x����"d "���myz������{�|}~����~��~��������~��SXTUXSTU�]UYU.Q�XSTU�]UYU
Tekniskt datablad Version 2011-01-10 Sikament ECO 12-2
1/1 Sikament ECO 12-21
Sikament ECO 12-2 Flyt/HRWR tillsatsmedel till betong
Användning Beskrivning
Sikament ECO 12-2 är ett mycket effektivt vattenreducerande flyttillsatsmedel av den tredje generationen som ger betongen god arbetbarhet och god styrke-utveckling. Sikament ECO 12-2 är CE- märkt i enlighet med: EN 934-2 och EN 934-6. CE-certifikat nr: 1274-CPD-704. Produkten uppfyller kriterierna för kemiska produkter i BASTA.
Användning Sikament ECO 12-2 kan användas till alla typer av betongkvalitéer inom betong-produktion där man vill uppnå enklare betonghantering, bättre och säkrare formutfyllnad, färre efterlagningar samt högre kvalitet och korttidshållfasthet.
Tekniska Data Färg och form Densitet pH-värde Kloridhalt Alkalieinnehåll, ekv Na2O Korrosionsegenskaper Torrhalt Viskositet Dosering
Röd vätska 1,05 kg/dm³ Ca 7 <0,10% av medlets vikt <0,3% av medlets vikt Icke relevant när bruksanvisning följs Ca 17% Lättflytande Ca 0,1-2,5% av cementvikten
Tillverkningsplats Sika Sverige AB Domnarvsgatan 15 163 08 Spånga SVERIGE
Förpackning Dunk 20 kg, fat 220 kg, transporttank 1100 kg samt tankbil. Lagringstid Minst 9 månader från leveransdatum (tankbil 16 månader). Förvaras frostfritt i
täckta kärl. Eventuell omröring skall ske med mekanisk alt. ”rundpumpning”. Farliga ämnen Se separat säkerhetsdatablad. Bruksanvisning
■ Automatisk doseringsutrustning för vikt- eller volymdosering rekommenderas, beakta densitet och torrhalt.
■ Tillsätt Sikament ECO 12-2 tidigast i samband med blandningsvattnet. ■ Genom att fördröja tillsättandet 20-30 sekunder ökar medlets effekt och kan
därigenom bättre utnyttjas. För optimal effekt kan tillsättningsordningen vara annorlunda än ovanstående, detta skall dock provas för varje enskilt fall.
■ Inverkan av överdosering kan förorsaka stenseparation vid flytkonsistenser och felpropotionerad betong. Liten retarderande effekt.
■ Sikament ECO 12-2 kan även kombineras med Sikas övriga tillsatsmedel, varvid varje medel tillsätts separat.
■ Rekommenderad blandningstid är 30-180 sek beroende på blandartyp. ■ Rekommenderad dosering är ca 0,1-1,8% av cementvikten. ■ Inverkan av detta tillsatsmedel kan variera beroende på vilket cement som
används. ■ Rengör utrustningen med vatten. ■ Förprov skall utföras med de aktuella delmaterialen till betongen enligt gällande
betongbestämmelser för klarläggande att avsedd effekt uppnås. ■ Använd plast, glasfiber eller rostfria tankar vid hantering av Sikament ECO 12-2.
Hälsa & Miljö Hälsa & Miljö
Se separat säkerhetsdatablad.
Lagstiftning Informationen och i synnerhet rekommendationerna avseende applikation och slutanvändning av Sikaprodukterna lämnas i god tro baserat på Sikas nuvarande kunskap och erfarenhet av produkterna när dessa lagras, hanteras och används under normala förhållanden på ett korrekt sätt. I praktiken kan differenserna i material, underlag och den aktuella platsen variera på sådant sätt att ingen garanti vad gäller användbarhet eller lämplighet för ett visst ändamål kan lämnas. Med hänsyn härtill kan något rättsligt ansvar av vad slag det må vara varken härledas från denna information eller från någon skriftlig rekommendation eller i övrigt beträffande produkten lämnade råd. Hänsyn måste vid användningen även tas till tredje mans ägande och andra eventuella rättigheter. Alla order accepteras under förutsättningen av att Sikas aktuella försäljnings- och leveransbestämmelser är gällande. Användaren skall alltid använda sig av den senaste utgåvan av den aktuella produktens tekniska datablad, vilket kan erhållas vid förfrågan eller på hemsidan www.sika.se.
Sika Sverige AB Domnarvsgatan 15 Box 8061 SE-163 08 Spånga Sverige
Tel. +46 8 621 89 00 Fax +46 8 621 89 89 www.sika.se
Tekniskt datablad Version 2008-08-05 Sika® UCS
1/1 Sika® UCS1
Sika® UCS Stabiliseringsmedel för undervattensbetong
Användning Beskrivning
Sika UCS är en betong- och brukstillsats speciellt framtagen för applicering under vatten. Den kan användas tillsammans med alla standardtyper av cement. Sika UCS är godkänd av Vattenfall Utveckling AB enl. BRO 2002:50. Produkten uppfyller kriterierna för kemiska produkter i BASTA.
Användningsområde Sika UCS ökar betongens nivellerande och självkomprimerande egenskaper. Den kan användas med Viscocrete, Sikament, Plastiment, SikaRetarder och andra Sika tillsatser.
Fördelar ■ Ökar sammanhållningen i betongen betydligt ■ Förlänger bearbetningstiden ■ Minimerar separation
Tekniska Data Färg och form Bulkdensitet Kloridhalt Lämpligt till Normaldosering Förpackning Lagring
Ljusgult pulver 0,33 kg/l <0,1 vikt-% Alla typer av portlandbaserade cementer inklusive SRC, PFA och GGBFS. 2 kg/m³ betong 5 kg hink 6 månader i torrt utrymme mellan +5ºC och +30ºC. Vid fuktig eller våt lagring uppstår klumpar och produkten kan inte användas.
Bruksanvisning Normala rekommendationer för blandning av undervattensbetong bör följas. Förprov av blandningens sammansättning, appliceringsprocess och dosering för varje projekt rekommenderas. Speciellt bör användandet av andra Sika tillsatser testas för att försäkra sig om att önskat resultat av utförandet har uppnåtts. Betong som innehåller Sika UCS bör pumpas till appliceringsstället. Under speciella förhållanden kan man låta betongen falla fritt genom vattnet utan onödiga separationsnivåer. Spill på våta ytor kan förorsaka halka. Torka upp eller skölj ner spill omedelbart. Dosering. Beroende på önskad effekt, typ av cement och blandningssamman-sättning är den normala doseringen 2 kg/m³ betong. Pulvret skall tillsättas under blandning/omrörning. Blandas noggrant. Blandningstiderna är normalt 50% längre än för en liknande blandning utan Sika UCS.
Hälsa & Miljö Hälsa & Miljö
Se separat säkerhetsdatablad.
Lagstiftning Informationen och i synnerhet rekommendationerna avseende applikation och slutanvändning av Sikaprodukterna lämnas i god tro baserat på Sikas nuvarande kunskap och erfarenhet av produkterna när dessa lagras, hanteras och används under normala förhållanden på ett korrekt sätt. I praktiken kan differenserna i material, underlag och den aktuella platsen variera på sådant sätt att ingen garanti vad gäller användbarhet eller lämplighet för ett visst ändamål kan lämnas. Med hänsyn härtill kan något rättsligt ansvar av vad slag det må vara varken härledas från denna information eller från någon skriftlig rekommendation eller i övrigt beträffande produkten lämnade råd. Hänsyn måste vid användningen även tas till tredje mans ägande och andra eventuella rättigheter. Alla order accepteras under förutsättningen av att Sikas aktuella försäljnings- och leveransbestämmelser är gällande. Användaren skall alltid använda sig av den senaste utgåvan av den aktuella produktens tekniska datablad, vilket kan erhållas vid förfrågan eller på hemsidan www.sika.se.
Sika Sverige AB Domnarvsgatan 15 Box 8061 SE-163 08 Spånga Sverige
Tel. +46 8 621 89 00 Fax +46 8 621 89 89 www.sika.se