Poly Vinyl Alcohol Fiber Reinforced Engineered Cementitious Composites

BUDAPEST UNIVERSITY OF TECHNOLOGY AND ECONOMICS DEPARTMENT OF CONSTRUCTION MATERIALS AND ENGINEERING GEOLOGY

Poly Vinyl Alcohol Fiber Reinforced Engineered Cementitious Composites

A jövő vasbetonja?

Pápay Zita Judit IV. evfolyam

Konzulens: Dr. Józsa Zsuzsanna

TDK Konferencia, 2004 november

2

Contents

1. Introduction 2. Review of the technical literature

2.1 Development of ECC 2.2 Theoretical background

3. Research work

3.1 Selection of contituents 3.1.1 Cements and fly ashes 3.1.2 Cements, fly ashes and sand

3.2 Mixing of ECC 3.2.1 Procedure of mixing

3.3 Strength, ductile behavior 3.3.1 Tensile Stress-Strain Representation 3.3.2 An for example: results of an experimental program in 2001

3.4 Our experimental program 3.4.1 Tensile tests

3.4.1.1 Preparations 3.4.1.2 The tests 3.4.1.3 The results

4. Classes of Target Applications

4.1 Durable infrastructure subjected to severe environmental loading 4.2 Infrastructure construction productivity

5. Cost of ECC

6. The ECC technology network

7. Future outlook

8. Conclusions

3

Poly Vinyl Alcohol Fiber Reinforced Engineered Cementitious Composites (PVA-ECC)

1. Introduction

During the last decade, a significant effort has been made developing a new class of

fiber reinforced cements (FRC), the so called engineered cementitious composites (ECC). This class of FRC exhibits tough, strain-hardening behavior in tension, despite containing low volumes of fibers. The superiority of ECC has been brought about by the engineering micromechanics approach and the development in fiber, matrix and processing technology. The importance of the interaction between the fibers and the matrix, governed by the fiber-matrix interface has been recognized, leading to interface modification techniques to engineer the desired properties. Fiber breakage is prevented and pull out from the matrix enabled instead, leading to tensile strain capacity in excess of 4% for ECC containing 2% by volume short Poly Vinyl Alcohol (PVA) fiber. PVA-ECC is a unique implementation of PVA fibers in a micromechanically designed matrix invented by Prof. Victor C. Li, professor of Civil and Environmental Engineering at the University of Michigan, Ann Arbor, and director of the Advanced Civil Engineering Materials Research Laboratory [1].

ECC is also an ultra-ductile fiber reinforced cement based composite, that has metal-like features when loaded is tension. The uniaxial stress-strain curve shows a „yield“ point, followed by strain-hardening up to several percent of strain, resulting in a material ductility of at least two orders of magnitude higher in comparison to normal concrete or standard fiber reinforced concrete. ECC has an unique cracking behavior. When loaded to beyond the elastic range, ECC provides crack width to below 100 µm, even when deformed to several percent tensile strain [2].

4

Fig.1.1 Strain-hardening means that PVA-ECC is considerably stronger in tension than any

regular or fiber-reinforced mortar1

Fig. 1.2 Strain-hardening and micro-cracking give PVA-ECC considerable ductility2

1 Source of the picture: http://www.kuraray-am.com/pvaf/pva-ecc.php 2 Source of the picture: http://www.kuraray-am.com/pvaf/pva-ecc.php

5

2. Review of the technical literature 2.1 Development of ECC

Since ECC was introduced about ten years ago, significant developments in research

and commercialization of ECC technologies have occurred both in the academic and in the industrial fields [3].

The effort to modify the brittle behavior of plain cement materials such as cement pastes, mortars and concretes has resulted in modern concepts of fiber reinforcement and interface engineering. In general, short fiber reinforced cement composites exhibit what is known as quasi-brittle behavior. This behavior is characterized by a more ductile post-peak softening in uniaxial tension compared with the plain matrix, as a result of gradual fiber pull-out from a single crack plane. It was demonstrated 10 years ago that a non-catastrophic failure mode exists when a brittle matrix is adequately reinforced either by continuous aligned fibers or short random fibers. The failure mode is characterized by a sustained or even higher load carrying capacity after first cracking of the matrix, as shown in Fig. 2.1.1 [4].

Fig. 2.1.1 Different tensile failure modes in cementitious composites [4]

The pseudo-strain-hardening behavior is associated with the appearance of a sequence of matrix cracks increasing in density until composite peak load is reached.

The short fiber reinforced composites designed to provide pseudo-strain-hardening properties based on micromechanical principles has been referred to as engineered cementitious composites (ECC). In the first generation of these composites have been used only cement and silica fume in their matrix, with no aggregates. The lack of fine and coarse aggregates in the matrix results in a low composite elastic modulus and may lead to high heat

6

of hydration, which may limit the widespread application of these composites in the construction industry.

In 1995 a new research began to design a new class of ECCs that can use a matrix incorporating suitable aggregates that will result in higher elastic modulus while maintaining the desirable feature of strain hardening. This object was achieved by the following procedure. First the effect of matrix mechanical properties on composite properties is reviewed from the micromechanical viewpoint with regard to conditions of composite pseudo-strain-hardening. A systematic experimental investigation is then conducted on the effect of matrix composition on matrix properties. It was demonstrated that only matrices with suitable fracture toughness, as defined by the micromechanical model, tend to retain the pseudo-strain-hardening property, while the composite elastic modulus is increased for all matrices with fine aggregates.

The figure 2.1.2 shows a flow-chart of some important elements of ECC R&D3, from basic material design theory to practical commercial applications.

Fig. 2.1.2 Flow-chart of important elements of the research and development of Engineered Cementitious Composites. Investigations into structural applications and commercial developments provide feedback to materials improvement via microstructure tailoring [3]

Micromechanics relates macroscopic properties to the microstructure of a composite, and forms the backbone of material design theory. Especially, it allows systematic microstructure tailoring of ECC as well as material optimization. Microstructure tailoring can lead to extreme composite ductility. An increasingly large database of mechanical (including tension, compression, shear, fatigue and creep) and physical properties (including shrinkage, and freeze-thaw durability) of ECC is now being established. Materials optimization also leads to compositions (e.g. moderately low fiber volume fraction less than 2 to 3%, coupled with suitable matrix design) that make it possible for very flexible materials processing. ECC can now be cast (including self-compacting casting), extruded or sprayed.

3 R&D= ’research and development’

7

Commercial development of ECC technologies imposes further considerations, in

addition to those described in the previous paragraphs. The adoption of a new technology must be justified, with advantages in cost-benefit ratio. While the initial raw material cost of ECC is higher than that of normal concrete, the long term benefits are sufficient to potentially drive this technology into the commercialization stage in the near future in a number of countries, including Japan, Korea, Australia, Switzerland and the US. Nonetheless, materials optimization for cost reduction of ECC remains important. It is expected that as investigation into structural response and commercial development of ECC technologies proceed, a feedback loop (Fig. 2.1.2) to microstructure tailoring for refinement in ECC materials will occur. In addition, special functionalities, such as lightweight structures, or high early strength, will drive the development of specialized versions of ECC for different applications with unique demands. Each of the R&D elements and especially the feedback process will benefit from collaborations among researchers, material suppliers, precasters, and design and construction contractors [3].

2.2 Theoretical background Extensive research has shown that the most fundamental property of a fiber reinforced

cementitious material is the fiber bridging property across a matrix crack, generally referred to as the s-d curve. This is the averaged tensile stress s transmitted across a crack with uniform crack opening d as envisioned in a uniaxial tensile specimen. The s-d curve provides a link between composite material constituents – fiber, matrix and interface, and the composite tensile ductility (Fig. 2.1.3).

Fig. 2.1.3 The linkages material constituents, crack bridging property and composite tensile ductility [3]

8

To understand the fundamental mechanisms governing strain-hardening ECC behavior versus tension-softening FRC behavior, it is necessary to recognize the load bearing and energy absorption roles of fiber bridging. The s-d curve (Fig. 2.1.4) can be thought of as a spring law describing the behavior of non-linear springs connecting the opposite surface of a crack, representing the averaged forces of the bridging fibers acting against the opening of the crack when the composite is loaded in tension.

Fig. 2.1.4 The s-d curve and the concept of complementary energy (shaded area labeled C). High bridging strength s cu and large complementary energy C are conducive to

composite strain-hardening [3]

One of the criteria for multiple cracking is that the matrix cracking stress (including

the first crack stress associated with the first crack) must not exceed the maximum bridging stress s cu . (The cracking stress is dominated by the matrix flaw size.) We may label this as the strength criterion for multiple cracking. A second criterion for multiple cracking is concerned with the mode of crack propagation, which in turn is governed by the energies of crack extension. We may label this as the energy criterion for multiple cracking. Clearly, violation of the strength criterion leads to a crossing-through crack which the loading cannot be supported by the fiber bridging stress. The energy criterion is less obvious and deserves a more detailed explanation.

When the fiber/matrix interface has too low strength, pull-out of fibers occurs, resulting in a s-d curve with low peak strength s cu . When the interface has too high strength, the springs cannot stretch, resulting in rupture and a small value of critical opening d

p . In either case, the complementary energy shown as the shaded area C to the left of the s-d curve in the figure 2.1.4 will be small.

9

Fig. 2.1.5 Steady state crack analysis [3] Steady state crack analysis reveals two crack propagation scenarios: in Fig. 2.1.5 (a)

above, low complementary energy in comparison to crack tip toughness results in Griffith type cracking. The fibers, illustrated schematically as springs, slide out or break in the mid-crack section where the opening d m exceeds d p in the s-d curve. In the figure 2.1.5 (b) high complementary energy results in a flat crack propagation configuration. The fiber/springs remain intact as the crack propagates under a constant steady state stress s ss , with the opening d m less than d p. After the forming of a flat crack, additional cracks can be formed under increasing load, resulting in the phenomenon of multiple cracking [3].

The shape of the s-d curve therefore plays a critical role in determining whether a composite strain-hardening as in ECC, or tension-softening, as in normal FRC, under uniaxial tensile load. The strength and energy criteria for multiple cracking provide guidelines for tailoring the fiber, matrix and interface for ECC materials. The shape of the s-d curve is governed by the fiber volume fraction, diameter, length, strength and modulus, in addition to the interfacial chemical and frictional bond properties. Controlling the shape of the s-d curve, therefore, boils down to controlling the fiber and fiber/matrix interaction parameters. This forms the tailoring strategy for the REC15 fiber now manufactured by Kuraray Co. Ltd. (Japan) for ECC reinforcement. The s-d curve has a direct bearing on the constitutive law in general, and the tensile stress-strain curve of the composite in particular, since it determines whether strain-hardening would occur or not. This composite material constitutive law in turn governs the response of a structure built with ECC material. Hence, the s-d curve may be considered as a critical link between materials design and structural design. From the above discussion, it can be seen that micromechanics, as already mentioned, serves as a useful tool to direct materials design for achieving desired structural performance.

Micromechanics also serves as a tool for materials optimization. The construction industry is a highly cost-sensitive industry. Any new material introduced must be cost-effective. Since fibers are much more expensive than cement, sand or water, it is imperative to minimize the amount of fibers used while maintaining the strain-hardening property. This concept is implemented in ECC in the form of minimizing the critical fiber volume fraction,

10

the amount of fibers that just switch the material from a normal tension-softening FRC behavior to a strain-hardening ECC behavior. The critical fiber volume fraction can be determined based on knowledge of fiber, matrix and interface properties. Using fiber content below this critical value will lead to normal FRC tension-softening behavior. On the other hand, using fiber content greatly in excess of this critical value leads not only high cost of material, but also creates difficulties in material processing. A fiber volume fraction at just above the critical value provides a composite optimal in performance, cost, and processability [3].

The parameters governing the behavior of short fiber reinforced cement composites can de divided into three groups. The first group consists of those related to fiber, such as fiber type, geometry and strength, etc., the second group consists of matrix related parameters such as tensile strength. Fracture toughness, elastic modulus, etc., and the third group consists of interface related parameters such as the interfacial bond strength. It is important to understand the quantitative influence of these parameters on composite properties in order to select a suitable type of matrix.

Based on the micromechanical theory of matrix crack extension and crack bridging by random discontinuous fibers, Prof. Li shows that the fiber content must exceed a certain critical volume fraction v for a brittle matrix composite to exhibit pseudo-strain-hardening under uniaxial tensile loading. This critical fiber volume fraction can be expressed in terms of fiber, matrix and interface parameters as follows:

v=12Jc /{ gt (Lf / df ) d 0 } (1)

where Jc is the composite crack tip toughness, and Lf and df are fiber length and diameter, respectively. The snubbing factor g and interface frictional bond strength t are the parameters which describe the interaction between fiber and matrix. The snubbing factor can be interpreted physically as the increase in bridging force across a matrix crack when a fiber is pulled out at an inclined angle, analogous to a flexible rope passing over a friction pulley. Finally, d 0 is the crack opening at which the fiber bridging stress reaches a maximum, s 0 , and is given by: d 0 = t Lf

2/Ef df(1+h) (2) where h =(Vf Ef)/(Vm Em), and Vf , Ef are the fiber volume fraction and elastic modulus, respectively, and Vm , Em are the matrix volume fraction and elastic modulus, respectively. When a brittle matrix shows pseudo-strain-hardening, the ultimate strength of the composite

s cu=1/2gt Vf (Lf / df ) (3) In Eq. (1), the composite crack tip toughness can be related to the matrix fracture toughness Km via Jc = Km

2(1-Vf )(1-nm2) / Em (4)

Where nm is the matrix Piosson`s ratio. The (1-Vf ) factor takes into account the reduced crack front dimension due to the presence of fibers. For our purpose, since the fiber volume fraction is limited to a few per cent, this correction can be negligible. For the same reason, the

11

term h in Eq. (2) can be negligible. Further simplification can be obtained by approximating the (1-nm

2) term as unity, so that Jc = Km

2/ Em

Fig. 2.1.6 shows the expected critical volume fraction v as a function of interface bond strength t, for different values of Jc [4].

Fig. 2.1.6 Influence of matrix fracture toughness and interfacial bond strength on critical fiber volume fraction (Ef =117 GPa, Lf =12.7 mm, df=0.038 mm, g=2.0, Em=25 GPa) [4]

12

3. Research work This research work was made at the Department of Technology and Testing of Building Materials of the Technical University Kaiserslautern ECC.

3.1 Selection of constituents The experiments that are mentioned in the last paragraph are rather complicated,

therefore we did not carry out them. We used empirical observations of foreign professors. The mixture consists of:

• Sand • Cement • Fly ash • Superplasticizer • Fibers

Types of cements:

• Cement I-32,5-R, Dyckerhoff Cement r=3,15 g/cm3 • Cement I-42,5-R-HS, Dyckerhoff Cement r=3,20 g/cm3 • Cement I-42,5-R-HS, Heidelberger Cement r=3,21 g/cm3 • Cement II / A-LL 42,5 R, Heidelberger Cement r=3,09 g/cm3

Types of fly ashes:

• SAFA SAFAMENT HKV r=2,31 g/cm3 • SAFA SAFAMENT SWF r=2,39 g/cm3

By the selection of the suitable constituents our guiding principle was to minimize the

water content of the mixture. For this reason we adopted the experiment of Okamura [8]. Determination of bp according to Okamura

After determination of materials’ densities all the components have to be correctly measured. We changed the water content proportionally to solid-state content (Vw/Vp). This proportion can be changed between 0.6 & 1.5. The labormortarmixer has to be set for the mixprogram according to EN 196-1 (60 s low revolution per minute, 30 s high revolution per minute, 90 s pause, 60s high revolution per minute). After mixing the slump flow should be measured. The measured can be accepted if it is between 140 and 230 mm. We make the evaluation according to the Eq. (5) below. A Gamma (G) value has to be determinated. G=(slump flow /1002)-1 (5)

13

In graphical representation the results are shown (G and Vw/Vp ) in coordinate system and they are connected with a line. bp is the point of intersection of this line and the vertical axis [8].

A summary of the results is presented in Table 3.1.1. First of all we examined the

cements and fly ashes separately, then according to the results we examined the mixtures of two different cements, two fly ashes with sand.

Table 3.1.2 Results of the cements or fly ashes

3.1.1 Cements

Constituent CEM I - 32,5 R, Dyckerhoff Cement

Water-solid state-volume proportion 1,32 1,37 1,42 1,47 Vw/Vp

solid state volume l 0,246 density kg/dm³ 3,15 weight g 775,0

water volume l 0,325 0,338 0,350 0,363 weight g 325,0 337,5 350,0 362,5 slump flow mm 165 171 186 198 rel. slump flow G 1,72 1,92 2,46 2,92 bp 1,13

Table 3.1.2 Results of CEM I - 32,5 R, Dyckerhoff Cement

Constituents CEM CEM + SWF

CEM + HKV

CEM+ SWF + QS

CEM+ HKV + QS

CEM I 32,5 R, Dyckerhoff Cement 1,13 1,01 0,95 CEM I 42,5 R - HS, Dyckerhoff Cement 1,25 1,06 0,95 CEM I 42,5 R - HS, Heidelberger Cement 1,20 0,67 0,68 CEM II / A - LL 42,5 R, Heidelberger Cement 1,28 0,96 1,01 0,71 0,73 SAFA SAFAMENT SWF 0,95 SAFA SAFAMENT HKV 0,62

14

Constituent CEM I - 42,5 R - HS, Dyckerhoff Cement Water-solid state-volume proportion 1,34 1,39 1,45 1,50

Vw/Vp solid state volume l 0,242

density kg/dm³ 3,20 weight g 775,0


Table 3.1.3 Results of CEM I - 42,5 R - HS, Dyckerhoff Cement

Constituent CEM I - 42,5 R - HS, Heidelberger Cement

Water-solid state-volume proportion 1,40 1,45 1,50 Vw/Vp

solid state volume l 0,300 density kg/dm³ 3,21 weight g 963,0

water volume l 0,420 0,435 0,450 weight g 420,0 435,0 450,0 slump flow mm 195 207 226 rel. slump flow G 2,80 3,28 4,11 bp 1,195

Table 3.1.4 Results of CEM I - 42,5 R - HS, Heidelberger Cement

Table 3.1.5 Results of CEM II / A -LL 42,5 R, Heidelberger Cement

Constituent CEM II / A -LL 42,5 R, Heidelberger Cement Water-solid state-volume proportion 1,37 1,42 1,47 1,52


density kg/dm³ 3,09 weight g 762,5


15

y = 0,1195x + 1,1275

y = 0,0995x + 1,2823y = 0,0622x + 1,252

y = 0,0749x + 1,1954

1,10

1,15

1,20

1,25

1,30

1,35

1,40

1,45

1,50

1,55

1,60

0,00 1,00 2,00 3,00 4,00 5,00

Rel. slump flow

Wat

er/s

olid

sta

te ra

tio

CEM I 32,5Cem I 42,5CEM I 42,5 Heidelberger CEMCEM IICEM I 32,5CEM IICEM I 42,5CEM I 42,5 Heidelberger CEM

Fig. 3.1.1 Graphical representation of the results of cements

In Cement I-42,5-R-HS, Heidelberger Cement the content of C3A was low,

therefore this cement needed lower water amount, than Cement I-42,5-R-HS Dyckerhoff Cement. Cement I-32,5-R-HS, Dyckerhoff Cemenet needs the lowest water amount, but we decided to use CEM I-42,5-R-HS, Heidelberger Cement and CEM II / A-42,5 R, that shows the third lowest value, because we also tried to optimize the mechanical properties.

3.1.2 Cements, fly ashes and sand According to the first part of the experiment, further on we used two of the four cements (CEM I-42,5-R-HS, Heidelberger Cement and CEM II / A-42,5 R). We made variations with both of the fly ashes, to determine, how the substances influence each other. We examined four different mixtures.

16

Constituent CEM II and FA-SWF and Sand mcem/mfa/msand = 1/1/1

Water-solid state-volume proportion 0,8 0,85 0,90 1,00


density kg/dm³ 2,70 weight g 811

water volume l 0,240 0,255 0,270 0,300 weight g 240,0 255,0 270,0 300,0 slump flow mm 130 154 181 224 rel. slump flow G 0,69 1,37 2,28 4,02

bp 0,7716

Table 3.1.6 Results of CEM II and FA-SWF and Sand

Constituent CEM II and FA-HKV and Sand mcem/mfa/msand = 1/1/1 Water-solid state-volume proportion 0,8 0,85 0,90



water volume l 0,240 0,255 0,270 weight g 240,0 255,0 270,0 slump flow mm 160 187 221 rel. slump flow G 2,50 3,88 bp 0,7375

Table 3.1.7 Results of CEM II and FA-HKV and Sand

17

Substances: CEM I 42,5 R, Heidelberger Cement and FA-SWF and sand mcem/mfa/msand = 1/1/1 Water-solid state-volume proportion 0,85 0,90 0,98




Table 3.1.7 CEM I 42,5 R, Heidelberger and FA-SWF and Sand

Substances: CEM I 42,5 R, Heidelberger Cement and FA-HKV and sand mcem/mfa/msand = 1/1/1

Water-solid state-volume proportion 0,75 0,80 0,85




Table 3.1.8 CEM I 42,5 R, Heidelberger Cement and FA-HKV and sand

18

y = 0,0592x + 0,7638

y = 0,0518x + 0,6553

y = 0,0425x + 0,7375

y = 0,0664x + 0,7133

0,6

0,65

0,7

0,75

0,8

0,85

0,9

0,95

1

1,05

0,00 1,00 2,00 3,00 4,00 5,00

Rel. slump flow

Wat

er/s

olid

sta

te ra

tioCEM II and SWF and sand

CEM II and HKV and sand

CEM I 42,5 R-HS, HeidelbergerCement and SWF and sandCEM I 42,5 R-HS, HeidelbergerCement and HKV and sandCEM II and SWF and sand

CEM I 42,5 R-HS, HeidelbergerCement and HKV and sandCEM II and HKV and sand

CEM I 42,5 R-HS, HeidelbergerCemeent and SWF and sand

Fig. 3.1.2 Graphical representation of the results of cements, fly ahses and sand As shown on the two diagrams (Fig. 3.1.1 and Fig. 3.1.2) the constituents

together behave in a different way. The lowest water demand is observed in the mixture of CEM I 42,5 R, Heidelberger Cement, HKV and sand. This means that the different substances have influence on each other, so we can only select the suitable constituents according to the water demands of the mixtures.

Our selected materials4:

• sand – fine grades with maximum size of 0.1 mm. These grades consist of rounded to sub-angular grains which provide high and

excellent permeabilities. • cement – Cement I 42,5 R-HS, Heidelberger Cement with low content of C3A • fly ash – SAFA SAFAMENT HKV • superplasticizer – Woermann Glenium ACE30 • fibers – Kuraray REC15 (specially developed for ECC)

4 For further information of the selected materials see appendix A-E.

19

We made photographs of the fibers with electron microscope to examine if there is an oiling agent on the surface of the fiber. It was demonstrated by a chemist at the department, that there is no special oiling agent on the surface of a fiber, nevertheless it is mentioned in the scientific articles about ECC.

Photo 3.1.1 Pictures of the electron microscope

Photo 3.1.2 Pictures of Japanese fibers with length of 12mm

20

3.2 Mixing of ECC

Quality control is very important before we start to mix. The appropriate weight of the correct constituent must be used, because very little difference makes considerable change in the consistency of the mixture. The procedure by which constituents is added is of importance. If this procedure is changed, our mixture will not show the properties of an ECC material.

3.2.1 Procedure of mixing

We should mix sand and cement approximately for 30-60s without any water, until the mixture becomes homogeneous. Then we add water (Photo 3.2.1). After that we add fly ash and superplasticizer (Photo 3.2.2) by turns, using SP only when the mixer can not mix further (Fig.3.2.3). At the end the fibers are added (Photo 3.2.4), but the mixture after that can only be mixed for 30 s! This 30 s has to be kept, if not, the mixture would be very clumpy (Photo 3.2.5).

Photo 3.2.1 Adding water Photo 3.2.2 Adding super

plasticizer

Photo 3.2.3 Mixture without fibers Photo 3.2.4 Adding fibers

21

Photo 3.2.5. After the mixing When the mixing is finished, the fiber dispersement should be quantified. We have to use our fingers to feel the mix, so we can state, that the mixture is:

1. very clumpy (Photo 3.2.6) 2. feeling of a few fibers in the mix (Photo 3.2.7) 3. very smooth – no fibers can be felt. (Photo 3.2.8)

Photo 3.2.6 The mixture is very clumpy Photo 3.2.7 A few fibers can be felt

Photo 3.2.8 The mixture is very smooth –

no fibers can be felt

22

3.3 Strength, ductile behavior

The unique feature of ECC is its ultra high ductility. This implies that structural failure by fracture is significantly less likely in comparison to normal concrete or FRC. In traditional R/C structural design, the most common and most important material parameter of concrete is its compressive strength. For this reason, structural strength (and more generally, structural performance) is often perceived to be governed by material strength. This means that higher material strength (usually referred to compressive strength in the concrete literature) is expected to lead to higher structural strength. This concept is correct only if the material strength property truly governs the failure mode. However, if tensile fracture failure occurs, a high strength material does not necessarily mean higher structural strength. Rather, a high toughness material, and in the extreme, a ductile material like an ECC, can lead to a higher structural strength [3]. 3.3.1 Tensile Stress-Strain Representation One of the most distinct features of ECC is that their inelastic tensile behavior can be represented by stress-strain relations. This is different from conventional random short-fiber-reinforced cement composites whose tensile behavior proceeds to the postpeak softening stage immediately after first cracking. The softening stage must be represented by a stress-crack opening-displacement relation as illustrated in Fig 3.3.1.

Fig. 3.3.1 Tensile Behavior Representation for Conventional random short-fiber-reinforced cement composites [6]

For the ECCs, tensile stress is maintained after first cracking, accompanying significant strain extension as depicted in Fig. 3.3.1.

23

Fig. 3.3.2 Tensile Behavior Representation for ECC [6] Fig. 3.3.2 illustrates that tensile behavior can apparently be expressed with a stress-strain relation up to the ultimate state, when damage localization initiates. The first crack state refers to the first bend-over point, when multiple cracking initiates. The ultimate state refers to the peak tensile stress state when multiple cracking terminates [6]. 3.3.2 An for example: results of an experimental program in 2001 In this research, an ECC was successfully developed with a tailored PVA fiber. Specifically, ECC containing 2 v% of a PVA-REC fiber with surface oil coating was demonstrated to have tensile strain capacity in excess of 4%. The results of the research illustrate the significance of micromechanics-based composite design for desirable performance. While PVA-ECC provides high performance, it remains economically feasible for practical engineering use because of its relatively low fiber content. For the same reason, an essentially standard mixing process for normal concrete can be applied to ECC, leading to a potentially broad range of applications in civil infrastructures [5].

Table 3.3.1 PVA-ECC tensile test results [5]

24

Fig. 3.3.4 Tensile stress versus strain curves of PVA-ECC (fiber with 0.3% oiling) [5]

25


26


27

Fig. 3.3.7 Crack width versus strain relationship [5]

Fig. 3.3.8 Multiple cracking of PVA-ECC: (a) fiber oiling agent 0.8%, v=2.0%, s/c=1.0); (b)

fiber oiling agent 0.8%, v=2.5%, s/c=1.2) [5]

28

3.4 Our experimental program We made a basic mixture according to the experiments of Okamura. Professor Fischer (Technical University Hawaii) helped us finding the correct proportions and improving our mixture. We made a basic mixture that is shown in Table 3.4.1.

Table 3.4.1Basic mixture

We expected that this mixture should have been the optimal solution and self-compacting. We varied the parameters as w/c-ratio, content of fibers, content of superplasticizer and content of stabilizer to examine how the results of the tensile tests change. The evaluation of these tests are now being made at the Technical University Kaiserslautern. In my present study I focus on the results of the basic mixture.

Density F (by weight) F (by volume) V [cm3] M-% m [kg] Cement (CEM I) 3,210 1 n/a 195 29 625

FA (HKV) 2,310 1 n/a 270 29 625 Sand 2,660 1 n/a 235 29 625 Water 1,000 0,42 n/a 262 12 262

SP (Glenium) 1,060 0,03 n/a 18 0,86 18,74 MC 1,300 0 n/a 0 0,00 0,00

Fiber PVA (12) 1,300 n/a 2,00% 20 1,19 26 kg Target vol. 1000 ml, liter 2181

W/C 0,42 W/(C+FA) 0,21 W/solids 0,14

(C+FA)/solids 0,67 density 2,18 kg/dm3

Date 23. Juli

Slump flow [mm] 265

29

3.4.1 Tensile tests

3.4.1.1 Preparations

We prepared specimens for tensile tests. We used glue to stick the specimens in steel

profiles (Fig. 3.4.1). The profiles were made directly for our experimental programme.

Photo 3.4.1 Preparation of a specimen Photo 3.4.2 Water curing The specimens were stored under water for minimum 21 days (Fig 3.4.2). Before the testing they were not dried or covered with any film layer.

30

3.4.1.2 The tests Specimens were stuck in the profiles (Fig. 3.4.3). We tested three specimens at the same time.

Fig. 3.4.3 Testing of the tensile strength

3.4.1.3 The results

type of the test tensile specimen marking 1 2 3 date of casting 27.7 27.7 27.7 date of testing 20.8 20.8 20.8 age of the specimen [d] 24 24 24 Clear span [mm] 120 120 120 length of the specimen [mm] 160 160 160 width of the specimen [mm] 38,3 38,8 38,9 height of the specimen [mm] 40 40 40,2 weigth [g] 525,1 543,3 536,1 volume [cm³] 245,1 248,3 250,2 density [g/cm³] 2,14 2,19 2,14 first crack force [kN] 5,9 4,5 4,7 ultimate crack force [kN] 4,2 4,3 3,2 surface of the cross-section [mm²] 1532,0 1552,0 1563,8 first crack strength [N/mm²] 3,9 2,9 3,0

mean value [N/mm²] 3,3 ultimate strength [N/mm²] 2,7 2,8 2,0

mean value [N/mm²] 2,5

Table 3.4.2 Test results

31

0

0,5

1

1,5

2

2,5

3

3,5

4

4,5

0,0 0,0 0,0 0,1 0,1 0,1 0,1

Strain [%]

Tens

ile S

tres

s [M

Pa]

Fig. 3.4.4 Tensile stress versus strain curve of specimen 1

0

0,5

1

1,5

2

2,5

3

3,5

4

0,00 0,05 0,10 0,15 0,20 0,25

Strain [%]

Tens

ile S

tres

s [M

Pa]


32

0

0,5

1

1,5

2

2,5

3

3,5

0,00 0,05 0,10 0,15 0,20 0,25

Strain [%]

Tens

ile S

tres

s [M

Pa]


The tensile-strain diagrams (Fig. 3.4.4, Fig. 3.4.5, Fig. 3.4.6) were prepared according

to the force versus deflection curves that can be found in the appendix F. As indicated in the tensile-strain diagrams the specimens did not show the properties of an ECC material. They should have showed minimum 3-5% strain capacity. This experiment was the first trial to make an ECC material in Germany, so the

mixture has to be optimized further. As I have mentioned the experimental program is now under evaluation. We varied the parameters as w/c-ratio, content of fibers, content of superplasticizer and content of stabilizer. On the other hand we have to develop the technology of mixing, storing, testing.

After the evaluation of the tensile tests of the basic mixture our first question was, if the specimens were really self-compacting or not. To answer this question we made a second series of this basic mixture with compacting.

33

0

1

2

3

4

5

6

0,00 0,05 0,10 0,15 0,20 0,25 0,30

Strain [%]

Tens

ile S

tres

s [M

Pa]

Fig. 3.4.7 Tensile stress versus strain curve of specimen 1 with compacting

0

1

2

3

4

5

6

0,00 0,02 0,04 0,06 0,08 0,10 0,12 0,14Strain [%]

Tens

ile s

tres

s [M

Pa]


34

0

1

2

3

4

5

6

0,00 0,02 0,04 0,06 0,08 0,10 0,12 0,14 0,16 0,18

Strain [%]

Tens

ile S

tres

s [M

Pa]


As shown in the diagrams the specimens with compacting were not more ductile, so we can exclude this problem. The already mentioned parameters (w/c-ratio, content of fibers, content of superplasticizer and content of stabilizer) and technology of mixing, storing, testing have to be optimized further to get a real ECC material. After 2 months of research, we got such a material, that behaves like a normal fiber reinforced cement. At the Technical University Kaiserslautern the research still continues, so it is likely to have better results soon.

35

4. Classes of target applications

Although FRC has been accepted by the practice community since the 1980s, most of its applications are limited to non-structural use, which can be at least partially attributed to the limited performance, economic constraint, lack of design guidelines, and inconvenient processing. Meanwhile, in constructed facilities, the need for improvement in structural ductility has received increasing attention. Designed for structural applications, ECCs bring extraordinarily high performance as an advanced construction material. Significant contributions to structural ductility, deformation capacity, strength, damage tolerance, and repairability have been demonstrated. Specifically, in steel-reinforced ECC (R/ECC), due to the strain-hardening behavior and large strain capacity, ECC can carry load well above the yield point of steel reinforcement. The high shear capacity of ECC helps to reduce or even eliminate the need for shear reinforcements. Small crack width limited to below 0.2 mm delivers superior durability, and the high ductility and compatible deformation of ECC with steel reinforcements lead to high structural damage tolerance and deformability.

Because of its moderate fiber content, ECC can be applied in on-site construction as

well as in off-site precast elements. Special versions of ECC with self-compacting rheological behavior have been developed. Extrusion of ECC pipes has been demonstrated. Broad application of PE-ECC, however, is hindered by the relative high cost of the PE fiber. Poly Vinyl Alcohol (PVA) fiber, which typically has a tensile strength between 1600 and 2500 MPa, was considered a promising alternative. The cost of PVA fiber is about 1/8 that of high-modulus PE fiber, and is even lower than that of steel fiber on an equal volume basis.

4.1. Durable infrastructure subjected to severe environmental loading

The most investigated applications in this category belong to repair of infrastructures using ECC. These include planned repairs of dam (Japan), the underdeck of a bridge (Japan), a sewage line (Korea), tunnel linings (Switzerland) and concrete bridge decks (US). There are a number of characteristics of ECC that make it attractive as a repair material. ECC can eliminate premature delamination or surface spalling in an ECC/concrete repaired system. Interface defects can be absorbed into the ECC layer, and arrested without forming spalls, thus extending the service life. Suthiwarapirak [3] showed that ECC has fatigue resistance significantly higher than that of commonly used repair materials such as polymer mortar. ECC also has good freeze-thaw resistance and restrained shrinkage crack control. There is increasing evidence that a crack width of 100 µm represents a threshold above which water flow through cracks becomes appreciable. Since the crack width of ECC can be kept below this limit, it is likely that ECC would serve as an excellent “concrete-cover” in R/ECC5 structures. The low transport rate of aggressive agents through ECC may delay steel corrosion leading to extended service life in, e. g. structures in coastal regions. Further research in needed to directly confirm this expectation.

4.2. Infrastructure construction productivity

ECC may lead to enhanced construction productivity in several ways. The most direct

means is by elimination of labor-intensive installation of shear reinforcing bars in seismic structures. The high shear capacity of ECC decreases or even eliminates the need for shear 5 R/ECC= ’steel fiber reinforced ECC’

36

reinforcement. In this topic several publications are available. The flexible processing routes of ECC also lend themselves to efficient methods of application of ECC in construction sites or in precast plants. For example, for many repair applications, or in tunnel lining construction, the use of spray processing can speed up the provide a continuous method of manufacturing high quality ECC products with low level of waste production. Self-compacting ECC lends itself to challenging construction conditions, including where horizontal formwork or “concrete” filled tubes are utilized, with potentially significant reduction in labor requirements.

5. Cost of ECC The additional cost of ECC over normal concrete derives mostly from the use of fibers

and higher cement content. In comparison to steel fibers used in many FRCs, polymer fibers such as PVA may be more expensive on a unit weight basis. However, it should be noted that polymer fibers have density six to seven times lower than that of steel, and it is the volume content of fibers and not the weight content which governs the performance of the cementitious composite. Partial substitution of cement with industrial by-products such as fly ash should further reduce the cost of ECC, although the resulting change in interface and matrix properties and their effects on composite strain capacity should be carefully examined. The economics of ECC should be based on cost/benefit analyses. The potential benefits of using ECC have been already discussed. The life cycle cost of a structure includes not only the initial material cost, but also the construction cost and maintenance cost. By reducing or eliminating shear reinforcement, there will be cost advantages in constructing infrastructure with ECC material, associated with the reduction of steel as well as reduction of on-site labor, while speeding up the construction process. The durability of ECC and ECC structures should extend the service life of infrastructures while reducing maintenance cost.

6. The ECC technology network

ECC is relatively new material emerging from laboratory to precast plants and construction sites. The material continues to evolve, and new material characteristics continue to be uncovered. More and more applications are being found for ECC. There is a need to exchange evolving information between academic, industrial and governmental concerns. For this reason, the ECC Technology Network was established in 2001. The ECC Technology Network is an informal organization composed of members that are interested in developing and promoting ECC technology. A web site hosted by the University of Michigan at www.engineeredcomposites.com provides an international platform for sharing news and knowledge of ECC materials and application technologies. There is no cost to joining this organization, just a commitment to further advance ECC. The current members are indicated in the map below (Fig. 5.1).

37

Fig. 5.1 The ECC Technology Network Members include academic, industrial and governmental organizations. They conduct research, development and commercialization of ECC technologies [3] 7. Future outlook

While rapid progress has been made in ECC technological development over the last decade, it may be expected that the coming decade will be even more exciting. As research advances, we will continue to discover more favorable characteristics of ECC that lend themselves to new infrastructure applications. It may be envisioned that a new generation of ECC material embodying the advantages of both steel (ductility) and concrete will be developed. These new materials will be

• designable for achieving targeted structural performance levels • sustainable with respect to social, economic and environmental

dimensions • self-healing when damaged • functional to meet requirements beyond structural capacity.

Associated with this material, a new generation of infrastructure system that have one or more of these characteristics will emerge

• safe with minimum repair needs even after subjected to severe loading conditions

• smart with self-adapting ability • mega-scale but without size-effect drawback • zero-maintenance even when exposed to severe environment • constructable at high speed and with low waste [3].

38

8. Conclusions At the Department of Technology and Testing of Building Materials of the Technical University Kaiserslautern ECC, a newly adopted concrete from the USA, is now being developed. I worked in this program for 2 months last summer (2004). ECC is an abbreviation for Engineered Cementitious Composites. In PVA-ECC, the mortar is optimized to work with PVA fiber at a low volume content. Physically and mechanically, ECC behaves like normal concrete. Under tension, however, ECC behaves more like a ductile metal. The strain capacity during strain-hardening (after first crack) is about 5%. This study surveys the development of ECC over the last decade since its invention in the early 1990’s, reviews the theoretical background of micromechanics based material design. I represent the details of my own experimental work in this project, which consists of the selection of the suitable substances and determination of the correct proportion of the materials in mixtures of an ECC. According to the results of the tensile stress tests of our basic mixture, the specimens do not show yet the properties of this new material. After 2 months of research, we got such a material, that behaves like a normal fiber reinforced cement, so the variable parameters (w/c-ratio, content of fibers, content of superplasticizer and content of stabilizer) of this basic mixture have to be optimized further. At the Technical University Kaiserslautern the research still continues, so it is likely to have better results soon.

39

For further information about ECC see Web-Sites:

1. Web-Site of Kuraray Specialities Europe GmbH Kuraray: http://www.kuraray-am.com 2. Web-Site of Prof. Li: http://ace-mrl.engin.umich.edu/NewFiles/director.html

3. Web-Site of Journal of Advanced Croncrete Technology (Japan Croncrete Institute): http://www.j-act.org/4-3.html

Web-Site of The ECC Technology Networks: www.engineeredcomposites.com

40

References [1] Goa Song and GPAG van Zijl, Tailoring ECC for commercial application [2] Victor C. Li, Durable Overlay Systems with Engineered Cementitious Composites (ECC), International Journal for Restoration of Buildings and Monuments Vol. 9, No. 2, pp. 1-20 (2003) [3] Victor C. Li, On Engineered Cementitious Composites (ECC) – A Review of the Material and Its Applications, Journal of Advanced Concrete Technology, Vol 1, No. 3, pp. 215-230 (2003) [4] Victor C. Li, Dhanada K. Mishra, Hwai-Chung Wu, Matrix design for pseudo-strain-hardening fiber reinforced cementitious composites, Materials and Structures, No. 28, pp. 586-595 (1995) [5] Victor C. Li, Shuxin Wang, Cynthia Wu, Tensile Strain-Hardening Behaviour of Polyvinyl Alcohol Engineered Cementitious Composite (PVA-ECC), ACI Materials Journal, No. 98, pp. 483-492 (2001) [6] Tetsushi Kanda, Zhong Lin and Victor C. Li, Tensile Stress Modeling of Pseudostrain Hardening Cementitious Composites, Journal of Materials in Civil Engineering, pp. 147-156 (May 2000) [7] DafStb-Richtlinie Selbstverdichtender Beton (SVB-Richtlinie), 12, Anhang P – Prüfverfahren zur Ermittlung des Wasseranspruchs, pp. 29-31 (August 2003)

41

Appendix A

Sand:

0

10

20

30

40

50

60

70

80

90

100

0,01 0,1 1

Aggregate size [mm]

Pass

ing

[%]

r=2,66 g/cm3

42

Appendix B Cement: from HeidelbergerCement Group

Source: http://www.heidelberger-zement.de/BTD.htm

43

Appendix C Flyash: from SAFA Saarfilterasche-Vertriebs-GmbH & Co. KG SAFA-FÜLLER® ist eine Steinkohlenflugasche die als Füllstoffbaustoff universell eingesetzt werden kann. Z.B. zur Verfüllung von tragenden Hohlräumen, nicht mehr genutzte Ver- und Entsorgungskanäle, Erdtanks oder Kavernen und für andere Anwendungen. Auf der Basis von Steinkohlenflugasche werden besondere Verfüllmassen hergestellt, die nach dem Aushärten hohe Festigkeiten erreichen und allen Sicherheitsansprüchen genügen. Gute Pumpbarkeit und die Neigung zur Selbstnivellierung werden durch die kugelige Form von SAFA-FÜLLER® erreicht. Schüttkegel- oder Luftblasenbildung ist deshalb nahezu ausgeschlossen.

Source http://www.safa.de/html/menu.html

44

Appendix D

Superplasticizer: from WOERMANN Bauchemie GmbH Glenium® ACE 30 (FM)

Artikel-Nr.:

0056 _________________________________________________________________

Empf. Dosierbereich in M-% vom Zementgewicht:

0,3 - 3,0

0,3 - 2,0 bei alkaliempfindlicher Gesteinskörnung _________________________________________________________________

Rohstoffbasis:

Polycarboxylatether _________________________________________________________________

Prüfungen und Zertifikate:

Fließmittel für Beton nach DIN EN 934-2: T 3.1 / 3.2 Entspricht DIN V 18998

Verwendung in Beton mit alkaliempfindlicher Gesteinskörnung entsprechend DIN V 20000-100, 8.1 (bei max. Dos. 2,0 %)

Verwendung in Beton mit alkaliempfindlicher Gesteinskörnung entsprechend DIN V 20000-100, 8.2 (Alkaligehalt ≤ 8,5 M-%)

Entspricht den Anforderungen der ZTV-ING und der ZTV-StB 01

Source: http://www.woermann.com/MBTWoermann_De/Home/default.htm

45

Appendix E

Fibers: from Kuraray Specialities Europe GmbH Fiber type

Diameter (mm)

Thickness (dtex)

Cut length(mm)*

Tensile Strength (N/mm)

Elongation(%)

Young's Modulus(kN/mm)

Specific Gravity Primary Applications

REC15 0.04 15 6,8,12 1600 (1.6GPa) 7 42 13 PVA-ECC and other

mortar reinforcement

Source: http://www.kuraray-kse.com

46

Poly Vinyl Alcohol Fiber Reinforced Engineered Cementitious Composites

Documents

development of ecc

mixing of ecc

cost of ecc

superiority of ecc

ecc technology network

fiberreinforced mortar1

fibermatrix interface

fiber breakage