ECOLOGICAL CONCRETE AND WORKABILITY: A MARRIAGE WITH FUTURE?

S.A.A.M. Fennis*, Delft University of Technology, the Netherlands J.C. Walraven, Delft University of Technology, the Netherlands
36th Conference on OUR WORLD IN CONCRETE & STRUCTURES: 14 - 16 August 2011, Singapore

Article Online Id: 100036007 The online version of this article can be found at: http://cipremier.com/100036007

Thisarticleisbroughttoyouwiththesupportof SingaporeConcreteInstitute www.scinst.org.sg AllRightsreservedforCIPremierPTELTD YouarenotAllowedtoredistributeorresalethearticleinanyformatwithoutwrittenapprovalof CIPremierPTELTD VisitOurWebsiteformoreinformation www.cipremier.com

36 Conference on Our World in Concrete & Structures Singapore, August 14-16, 2011

th

ECOLOGICAL CONCRETE AND WORKABILITY: A MARRIAGE WITH FUTURE?

S.A.A.M. Fennis* and J.C. Walraven Civil Engineering and Geosciences Delft University of Technology Stevinweg 1, 2628 CN, Delft, the Netherlands e-mail: s.a.a.m.fennis@tudelft.nl, webpage: www.concrete.citg.tudelft.nl
*

Keywords: Ecological concrete, particle packing, rheology,

Abstract. Optimizing concrete mixtures with regard to cement content is one of the most important factors in the design of ecological concrete. By minimizing the cement content, energy consumption and CO2 emissions of concrete are reduced, and residual products from other industries are reused. By making use of particle packing optimization techniques, it is possible to lower the cement content in concrete without changing concrete properties in a negative way. The high density of the particle structure leaves less space for voids to be filled with water, which reduces the water demand and increases the strength of concrete. To be able to predict how much water is necessary to fill up the voids in between the particles of a concrete mixture and predict the workability, the particle packing density of the materials should be known. Particle packing density of powders and mixtures was tested by the mixing energy method. Twenty cement pastes and more than hundred mortar mixtures were tested to relate the packing density to the workability of mixtures. Based on the experience with the cement pastes and mortars, three ecological concrete mixtures with low cement contents were designed. This was done following the cyclic design procedure presented in this paper. The relations between packing density and water demand and between particle structure and the cement spacing, are used in this design procedure to predict the strength of ecological concrete. With this method, the amount of water in concrete mixtures can be reduced. The reduced amount of water and low water/cement ratio can be used to save a certain amount of cement when concrete is designed for a fixed strength class. In this way, the cement content in concrete can be reduced from the prescribed minimum of 260 kg/m3 in ordinary concrete to 110 kg/m3 in ecological concrete. Experiments have shown that with this 57% reduction of Portland cement, material properties such as strength and modulus of elasticity can remain constant or even improve. However, it is important to control the water demand of these mixtures. A slight increase in water demand in mixtures with a low cement content has a larger effect on the water/cement ratio than in mixtures with a high cement content. Thus, in the future it will be very important to use particle packing models to control the water demand and workability of ecological concrete.

_______________________________________

Delft University of Technology

S.A.A.M. Fennis and J.C. Walraven

INTRODUCTION

Ecological concrete can be created in several ways, for instance by optimizing the concrete mixture, by minimizing material transports or by using recycled materials. However, about 50% of the total CO2 emitted during the construction of a concrete structure comes from the use of Portland cement. Therefore, optimizing concrete mixtures with regard to cement content is one of the most important factors in the design of ecological concrete. By minimizing the cement content, not only energy consumption and CO 2 emissions of concrete are reduced, but also residual products from other industries, like fly ash or incinerator bottom ash, are reused. Therefore, less material is dumped as landfill and more natural resources are saved. By making use of particle packing optimization techniques, it is possible to lower the cement content in 1 concrete without changing concrete properties in a negative way . This is done by optimizing the concrete composition in such a way that the highest packing density is achieved. The optimization results in a stiff and strong particle structure, which has a positive influence on the mechanical properties such as shrinkage and creep. Furthermore, the high density of the particle structure leaves less space for voids to be filled with water, which reduces the water demand and increases the strength of concrete. The water demand of the particle structure is very important since a slight increase in water demand in mixtures with low cement content results in higher water/cement ratios. A higher water/cement ratio decreases strength and durability. Therefore, the key to design ecological concrete is to control the water demand and workability. In this paper it is presented in what way particle packing density and water demand are related and how this can be used to design concrete with lower cement contents.

PARTICLE PACKING DENSITY AND WATER DEMAND

The particle packing density of an ecological concrete mixture has an important influence on its water demand. The definition of particle packing density is the solid volume of particles in a unit volume. However, a distinction should be made between a stable particle structure and the solid content mix of a real concrete mixture. In a stable particle structure all particles are in contact with each other. In such a stable structure, the particles are packed with a certain packing density t. This packing density is higher than the solid content mix of a real concrete mixture containing that particle structure. This is shown in Figure 1, where the same amount of particles in a stable particle structure (left hand side) is packed closer than in a real mixture (right hand side). In a real concrete mixture part of the water is used to fill the voids between the particles, while the rest of the water is regarded as excess water. This excess water provides the flowability of the mixture. Flowability increases with higher amounts of excess water in the mixture. In that case, the solid content of the mixture mix decreases.

Excess water

Stable parcle structure

Mixture

Figure 1: The volume occupied by a stable particle structure (left hand side) compared to the volume occupied by a flowable mixture consisting of that particles (right hand side)

S.A.A.M. Fennis and J.C. Walraven

Increasing the packing density improves the workability of a mixture, which is very useful in the design of ecological concrete mixtures. If the particles of a concrete mixture are optimized in such a way that the particle packing density increases, less void water is necessary. If in this optimized mixture the water dosage would be kept constant, the water that first filled up the voids between the particles now becomes available as excess water and provides more flowability.

TESTING METHODS

To be able to predict how much water is required to fill up the voids in between the particles of a concrete mixture and predict the workability, the particle packing density of the materials should be known. The maximum packing density of dry particles can be determined according to NEN-EN 10973:1998 for loose bulk density. The method can be extended to determine the maximum packing density at a certain compaction level, by applying external loads such as vibration or top-weight. However, with fine particles (< 1 mm), the inter-particle forces become increasingly important. These inter-particle forces can cause, for instance, agglomeration of particles, thus lowering the packing density. Since the inter-particle forces depend on the conditions (dry, wet, use of superplasticizer etc.) of the particle structure, also packing density is influenced by this. Therefore, it is important to measure the maximum packing density of the particles under the same conditions as under which the particles would be used in concrete. However, to determine the maximum packing density of wet particles, no single method is generally accepted and therefore different countries use their own test methods to determine packing density and/or water demand of fine particles. The following list gives a short summary of the most frequently used techniques: 2 Water demand France 3 Water demand Germany 4 Water demand - mixing energy Proctor test (NEN-EN 13286-2) 5 Centrifugal consolidation 6 Water demand Japan 7 Rheology Krieger and Dougherty 7 (Gas) pressure filtration 8 Vicat test 9,10 Shear compactor For the experiments presented in this paper, the mixing energy method was used to determine the packing density of the powder. Each mixture consisted of 1500 grams of powder. The mixing procedure started by mixing the dry powders with (an estimated) 95% of the water and the superplasticizer (1.2% kg/kg Glenium 51) in a three-liter Hobart mixer for one minute. After two minutes resting and scraping, mixing was continued with a constant water supply of 0.3 ml/s until maximum mixing power consumption was reached in about 2 minutes. Tests on each powder type or size fraction were performed at least twice to ensure a packing density precision of 0.002 [-]. The average amount of water in the mixture at maximum power consumption was taken as the void volume between the particles to calculate the packing density. To relate the packing density to the workability of mixtures, tests were performed on cement pastes and mortar mixtures. The pastes were mixed in a three-liter Hobart mixer. First, the dry powders were mixed for ten seconds after which the water and superplasticizer were added. The paste was mixed for 1 minute at low speed, then rested and scraped for one minute and subsequently mixed for another minute at low speed. The slump flow was determined by a mini cone test (upper/lower diameter 20/37 mm and height 57 mm) on a flow table (Tonindustrie) with a 300 mm diameter glass plate. The slump flow was taken as the average spread diameter, calculated in four directions. A coaxial cylinder viscometer, PAAR Physica MC1, was used to determine the viscosity of the paste. The mortars were mixed in a three-litre Hobart mixer following the mixing procedure of NEN-EN 1961:2005. The slump flow was determined by a mini cone test (upper/lower diameter 38/90 mm and height 75 mm) on a flow table (Tonindustrie) with a 300 mm diameter glass plate.

S.A.A.M. Fennis and J.C. Walraven

Furthermore, concrete mixtures were tested on strength development in time. Cube compressive strength and tensile splitting strength were measured according to NEN-EN 12390:2009 on 150 mm cubes after 2, 7, 28, 56 and 90 days. At every point in time three specimens for compressive strength and three specimens for splitting tensile strength were tested. Additionally, prism compressive strength and modulus of elasticity were determined after 28 days. The experiment was executed by loading three 100100400 mm prisms with a -3 constant deformation speed of 10 mm/s. The longitudinal displacements were recorded continuously by means of four LVDTs per specimen, which were placed in longitudinal direction in the center of the long sides with 136 mm measuring length.

MATERIALS AND MIXTURE DESIGN

The maximum packing density of seventeen pastes was tested by the mixing energy method. The mixtures consisted of cement (CEM I 42.5 N ENCI Maastricht), quartz powder (M6, M300, M600 Sibelco) or a combination of both. 10, 20, 30 or 40% [kg/kg] of the total cement volume was replaced by quartz powder. Quartz powder from three different size classes was used (M6: d50=33m, M300: d50=14m, M600: d50=3m). To minimize the differences in mixing energy between the mixtures, all mixtures consisted of 1500 grams of powder to which the water and superplasticizer (Glenium 51 BASF con.35%) were added according to the procedure described in the testing methods. For mixture compositions see Table 1. For the mortar tests, the same powder compositions were tested according to NEN-EN 1961:2005, including the same amount of superplasticizer (1.2% per gram cement) and with an adjusted water content to ensure stability of the mixtures. The mixture compositions are presented in Table 2. The mortar mixtures were tested on flowability and cube compressive strength at 7 and 28 days. Based on the experience with the cement pastes and mortars, three ecological concrete mixtures with low cement contents were designed. The mixtures were designed for strength class Table 1: Mixture compositions and test results of cement pastes series A.
CEM I 42.5 N [g] A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 A11 A12 A13 A14 A15 A16 A17 A18 A19 A20 CP100_W026 CP90-M6 CP80-M6 CP70-M6 CP60-M6 CP90-M300 CP80-M300 CP70-M300 CP60-M300 CP90-M600 CP80-M600 CP70-M600 CP60-M600 CP100_W029 CP100_W033 CP100_W037 CP100_W043 M6 M300 M600 1500 1350 1200 1050 900 1350 1200 1050 900 1350 1200 1050 900 1500 1500 1500 1500 Quartz powder M6 [g] 0 150 300 450 600 0 0 0 0 0 0 0 0 0 0 0 0 1500 M300 [g] 0 0 0 0 0 150 300 450 600 0 0 0 0 0 0 0 0 1500 M600 [g] 0 0 0 0 0 0 0 0 0 150 300 450 600 0 0 0 0 1500 Water [g] 378.3 378.3 378.3 378.3 378.3 378.3 378.3 378.3 378.3 378.3 378.3 378.3 378.3 421.8 475.8 545.4 637.8 393.2 466.3 496.4 Glenium 51 [g] 18.0 18.0 18.0 18.0 18.0 18.0 18.0 18.0 18.0 18.0 18.0 18.0 18.0 18.0 18.0 18.0 18.0 WCR [-] 0.26 0.29 0.33 0.37 0.43 0.29 0.33 0.37 0.43 0.29 0.33 0.37 0.43 0.29 0.33 0.37 0.43 WPR [-] 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.26 0.29 0.33 0.37 0.43 e t CIPM Viscosity [Ns/m ] 1.23 1.37 1.19 1.33 1.22 1.50 1.28 1.27 1.35 1.33 0.95 0.67 0.45 0.55 0.27
** **
2

Kt=12.2 [-] [-] 0.604 0.606 0.61 0.616 0.614 0.602 0.605 0.601 0.603 0.649 0.669 0.669 0.666 0.604 0.604 0.604 0.604 0.583 0.542 0.527* 0.604 0.602 0.600 0.598 0.596 0.597 0.589 0.582 0.575 0.637 0.660 0.656 0.639 0.604 0.604 0.604 0.604 0.583 0.542 0.527

* Due to required high mixing energy, one measurement only ** below measuring range

S.A.A.M. Fennis and J.C. Walraven

C20/25 with minimum amounts of cement CEM I 42.5 N (ENCI Maastricht). Three types of cement replacing materials were used: fly ash because of its ball bearing effect and high packing density, quartz powder because of its fine and narrow particle size distribution, and ground incinerator bottom ash as residual product with low environmental impact. The mixture compositions of the three ecological mixtures and their reference mixture are presented in Table 3 3 as series C. The reference mixture C1 was designed with 260 kg cement per m concrete, which is the minimum cement content according to the Dutch standards. Since the mixture was 2 designed for strength class C20/25 with design strength 33 N/mm , a relatively high amount of water had to be dosed. Mixtures C2 and C3 were optimized on packing density. Using the design 3 cycle as presented in section 6, the cement content was minimized to 110 kg/m in order to reach 2 the estimated design strength of 33 N/mm . Mixture C2 contained a combination of fly ash and quartz powder M600. In mixture C3 blast furnace slag cement CEM III/B 42.5 N was used to increase the packing density even further and create a more ecological mixture containing less Portland clinker. Mixture C4 was designed with wet ground municipal solid waste incinerator bottom ash from INASHCO. 5 RESULTS

Replacing part of the cement by the coarse filler M6 slightly improves the packing density of the pastes, see Table 1. When cement is partly replaced by M300, the experimentally determined packing density e is decreased a little. The addition of M600 shows a significant improvement of the packing density. Replacing 20% or 30% of cement by M600 can increase the packing density by more than 10%. The relatively small differences in the particle packing density of the pastes containing M6 and M300 do not show a clear relation with the viscosity measurements. Adding M600 to the cement paste clearly shows an increase in the packing density as well as a decrease of the viscosity. However, the number of measurements was too low to demonstrate a direct relation between packing density and viscosity. The flow measurements of the mortar mixtures show the same trends as the paste mixtures. Replacing cement by the coarser quartz powders M6 and M300 has less influence on the packing density and flowability of the mortar mixtures than using the fine quartz powder M600. The increased packing density and increased flowability that are the result of replacing cement with M600, can be used to optimize mixtures. This was done for mixtures B14 and B15 which were designed to have the same flowability as the reference mixture B1. Because of the increased packing density these mixtures require less water to fill the voids. Therefore, the required amount of water at this consistency level could be reduced by more than 20%. Table 2: Mixture compositions and test results of mortars series B.
CEM I 42.5 N [g] B1 B2 B3 B4 B5 B6 B7 B8 B9 B10 B11 B12 B13 B14 B15 C100 C90-M6 C80-M6 C70-M6 C60-M6 C90-M300 C80-M300 C70-M300 C60-M300 C90-M600 C80-M600 C70-M600 C60-M600 900 810 720 630 540 810 720 630 540 810 720 630 540 Quartz Sand Water powder [g] 90 180 270 360 90 180 270 360 90 180 270 360 [g] 2700 2700 2700 2700 2700 2700 2700 2700 2700 2700 2700 2700 2700 [g] 326 326 326 326 326 326 326 326 326 326 326 326 326 SP 51 [g] 10.8 10.8 10.8 10.8 10.8 10.8 10.8 10.8 10.8 10.8 10.8 10.8 10.8 WCR [-] 0.37 0.41 0.46 0.53 0.62 0.41 0.46 0.53 0.62 0.41 0.46 0.53 0.62 0.37 0.37 WPR [-] 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 0.37 t Slump Slump Kt=9 [-] 0.803 0.801 0.798 0.796 0.793 0.801 0.798 0.795 0.792 0.813 0.821 0.827 0.829 [mm] 2 5 5 2 12 3 3 13 16 15 27 39 48 4 2 Flow Comp. strength flow value 7-day 28 day 2 2 [mm] [mm] [N/mm ] [N/mm ] 83 91 83 84 89 86 87 92 93 90 92 134 159 86 85 95 95 95 91 102 98 97 124 113 97 140 177 202 96 94 49.6 42.8 36.8 33.6 24.1 47.6 37.7 35.6 27.8 47.8 47 38.9 35.9 54.9 57.6 65.5 59.6 49.0 42.0 34.7 61.6 48.6 46.6 38.5 62.6 60.4 52.6 50.0 68.0 74.4

C90-M600 810 C80-M600 720

90 2700 180 2700

293 10.8 259 10.8

0.33 0.813 0.30 0.821

S.A.A.M. Fennis and J.C. Walraven

Further investigations were performed on more than hundred mortar mixtures , showing a relation between flow value and the packing density of a mixture. The results are presented in Figure 2 as a function of the solid volume in a real mixture mix over the maximum calculated packing density t.
250 200

11

Flow value [mm]

150 100 R = 0.87 50 Mortar mixtures 0 0.85 0.9 0.95 1

11

mix/t [m3]

Figure 2: The flow value as a function of mix/t for mortar mixtures

The ecological concrete mixtures were designed based on the results from the mortar tests. Increasing the packing density by using the right type of cement replacing material decreases the water demand. In combination with the binding effect of the fly ash and bottom ash, mixtures 3 could be designed for the same strength class C20/25 with only 110-125 kg/m cement. Evaluation of the fresh state of these ecological concrete mixtures showed that all mixtures in series C were homogeneous and stable. Slump measurements (NEN-EN 12350-2:2009), air content (NEN-EN 12350-7:2009) and density (NEN-EN 12350-6:2009) are presented in Table 3. All the mixtures had sufficient workability for casting and vibration. Mixture C3 was judged as harsh compared to the other mixtures, which was especially noted when filling the moulds. To increase the ability to cast in practice, it is recommended to increase the total powder content. This will also improve the cohesion of the mixtures which was evaluated as low for mixtures C3 and C4, leading to easy collapse of the slump. These two mixtures had similar workability as mixture C2, namely consistency class C1 earth moist. Very low segregation and little bleeding of the mixtures was observed, by virtue of the dense particle structure. The cube compressive strength development of series C is presented in Table 3. All mixtures 2 reached the mean strength of 33 N/mm for strength class C20/25 within 28 days. The contribution of the fly ash and quartz powder to decrease the water demand is substantial. However, neither the conventional water/cement ratio nor the water/binder ratio of mixtures C2 and C3 can explain the strengths attained by these mixtures. Mixture C2 has a water/cement ratio of 0.94 and a water/binder ratio of 0.83 (NEN-EN 206-1:2001). Mixture C3 has a water/cement ratio of 0.94 and a water/binder ratio of 0.89. Based on the water/binder ratio the 2 28-day strength of these mixtures should have been between 20 and 25 N/mm . The additional 2 2 strength gain up to 39.6 N/mm for mixture C2 and up to 33.5 N/mm for mixture C3 is explained by the high particle packing density of the powders and possible additional pozzolanic effects of the fly ash and quartz powder, both improving the microstructure of the concrete. This is also confirmed by electrical resistance measurements. Mixture C4 contained ground IBA as cement 2 replacing material and reached a 28-day cube compressive strength of 37.9 N/mm . The relationship between the average cube compressive strength and the tensile splitting strength of the ecological mixtures is the same as for normal concrete. Also the measured moduli of elasticity of the ecological mixtures comply with the relation for compressive strength versus modulus of elasticity as described in Eurocode 2 for normal concrete.

S.A.A.M. Fennis and J.C. Walraven

Table 3: Mixture compositions [kg/m ] and the measured material properties of series C. Composition CEM I 42.5 N (ENCI) CEM III/B 42.5 N (ENCI) Fly ash (SMZ Maasvlakte) Quartz powder M600 (SIBELCO) Ground IBA (INASHCO) Aggregates 4-16 (Filcom) Sand 0-4 (Filcom) Glenium 51 [% kg/kg of powders] Effective amount of water Water/cement ratio Water/powder ratio Estimated density (1% air) Packing density CIPM Rheological properties Slump Air content Density Mechanical properties 2-day cube compressive strength 7-day cube compressive strength 28-day cube compressive strength 56-day cube compressive strength 90-day cube compressive strength 7-day tensile splitting strength 28-day tensile splitting strength 56-day tensile splitting strength 90-day tensile splitting strength 28-day prism compressive strength 28-day modulus of elasticity [kg/m ] 3 [kg/m ] 3 [kg/m ] 3 [kg/m ] 3 [kg/m ] 3 [kg/m ] 3 [kg/m ] [%] 3 [kg/m ] [-] [-] 3 [kg/m ] [-] [cm] [%] 3 [kg/m ] [N/mm ] 2 [N/mm ] 2 [N/mm ] 2 [N/mm ] 2 [N/mm ] 2 [N/mm ] 2 [N/mm ] 2 [N/mm ] 2 [N/mm ] 2 [N/mm ] 2 [N/mm ]
2 3

C1 260 1193 718 0.8 162 0.62 0.62 2351 0.886 20 0.9 2366 13.9 24.2 32.1 36.4 36.4 2.0 2.5 2.6 2.7 29.5 30500

C2

Mixtures C3 44 66 65 85 1160 866 1.2 103 0.94 0.40 2408 0.898 12 0.9 2424 7.2 17.6 33.5 35.3 39.1 1.4 2.5 3.0 2.6 20.0 30500

C4 125 75 50 1157 864 1.2 112 0.90 0.45 2402 0.890 14 0.9 2456 12.9 22.9 37.9 48.0 55.1 2.0 3.0 3.3 4.0 29.1 30500

110 88 62 1162 867 0.8 103 0.94 0.40 2409 0.897 3 1.8 2406 15.2 25.2 39.6 39.9 53.1 2.1 2.7 2.7 3.7 26.6 32500

DESIGN METHOD FOR ECOLOGICAL CONCRETE

In the mixture design of ecological concrete, increasing the packing density influences the concrete properties in different ways. Increasing the packing density reduces the water demand and affects the flowability of concrete. Furthermore, with an optimized packing density the particles in the structure are closer to each other, which affects concrete properties such as strength and shrinkage. These two effects can be used in the design of ecological concrete. The relation between packing density and water demand can be used to predict the required amount of water of a mixture. As described in section 2, the water demand of a mixture depends on the required amount of void and excess water. Increasing the packing density lowers the required amount of void water. For instance, in mixtures B1 and B10-B13 packing density increased and less void water was required. Since the amount of water in the mixtures was kept constant, the water became available as excess water and the flowability of these mixtures increased. The other way around, mixtures B14 and B15 were designed to have the same flowability as the reference mixture B1. Because B14 and B15 have a higher packing density, they require less (void) water for this flowability. Therefore, the mixtures could be designed with less water and the possibility was created to maintain a constant water/cement ratio while replacing cement.

S.A.A.M. Fennis and J.C. Walraven

Optimizing the packing density is not only beneficial for the water demand, but it can also decrease the space between the cement particles. This space between the cement particles depends on the particle structure of the concrete. In ecological concrete this factor is important because the cement is replaced by fillers (or binders). Depending on the particle size of the used filler, the space between the cement particles will vary, as is shown in Figure 3. Replacing cement by coarse fillers, such as M6 and M300, increases the space between the cement particles. This can be explained by considering a container filled by cement particles. If 20% of the cement is replaced by a filler of similar size, the packing density remains constant. In total the same volume percentage of the container is still occupied by the particles, however, now only 80% of the original cement particles are present in that container. This means that the cement particles are relatively further away from each other (see also Figure 3 left hand side). However, when cement is replaced by finer particles, such as M600, the packing density increases. In that case, the 80% cement particles can be present in a smaller container (see Figure 3 right hand side) and therefore the space between the cement particles is smaller than with the coarse filler.
C

Figure 3: The volume occupied by a stable particle structure with coarse filler (F) compared to the volume occupied by a stable particle structure with fine filler (C=Cement). The relations between packing density and water demand and between particle structure and the cement spacing, can be used in the design of ecological concrete. Figure 4 presents the cyclic design procedure for ecological concrete based on these two relations. The cyclic design procedure starts by determining the packing density t of the particles in a concrete mixture. It 2,12-14 can be calculated by analytical particle packing models . For this research project on 2 ecological concrete the Compressible Packing Model was extended to the Compaction 11 Interaction Packing Model (CIPM) to improve its performance for fine fillers . The CIPM was developed and tested to include wet packings of microparticles in combination with superplasticizer Glenium 51. The result of this first design step is a known particle size distribution with its corresponding packing density. In the next step the water demand of the mixture is determined by using of mix/t, see also Figure 2. This value is directly related to the flowability of a mixture, so mix and thus the volume of water can be adjusted to the required consistency. When water absorption of the materials and predicted air content are available, the amount of water to be dosed during the mixing process is known and thus the mixture composition is available. In the following step the strength is predicted based on this mixture composition, its packing density and its particle structure. The strength prediction is based on the assumption that in low strength ecological concrete all aggregates are stronger than the produced concrete. In that case, the cement glues the aggregates together and the strength depends on the space to be bridged to connect all aggregates. With a high packing density in the mixture, cement particles and other particles are close to each other, reducing the space that needs to be filled by hydration products. This leads to higher strengths. With higher amounts of water and higher water/cement ratios, the cement particles are further apart. Therefore, during the hydration process, the hydration products of the cement particles need to bridge a larger distance, eventually leading to lower strengths. By using 11 this concept of cement spacing factor, the strength of the mixture can be predicted .

S.A.A.M. Fennis and J.C. Walraven

packing
xtu Mi

mi / x t

re ju ad

en stm t

water

CSF

strength

Figure 4: Cyclic design procedure for ecological concrete. After the strength prediction, adjustments can be made to the mixture composition to include requirements of the user. For instance, when the strength is still higher than required, cement can be replaced by a filler in order to create a more ecological mixture. After changing the mixture composition, particle size distribution changes and the cyclic design procedure should be followed again: calculation of the new packing density, determination of the water demand and prediction of the strength. This cyclic design is repeated until the mixture meets all requirements set by the user. For this project on ecological concrete the only controlling parameter in the design procedure was set to be the strength requirement. However, the set-up of the cyclic procedure allows also for concrete mixture design based on mixture composition restrictions, or mixture design based on defined performance requirements.

CONCLUSIONS AND FUTURE PROSPECTS

The workability of ecological concrete is very important when designing these types of concrete mixtures. The flow value is related to the packing density and the amount of water in the mixture. Higher packing densities decrease the void volume between the particles and therefore decrease the amount of water necessary to fill this void volume. In a real mixture, part of the added amount of water is used to fill voids, while the rest is used to lubricate particles and allow flowability. It is important to control this amount of water because the excess water pushes the cement particles further away from each other, thus increasing the volumetric distance between the cement particles. In that case, a higher water/cement ratio and lower solid volume lead to larger distances between the cement particles and thus, eventually, to a lower strength due to the fact that the hydration products of the cement particles need to bridge larger distances. In ecological concrete with high amounts of cement replacing materials, this effect is even more important than in ordinary concrete. This is because a slight increase in water demand of mixtures with a low cement content has a larger effect on the water/cement ratio than in mixtures with a high cement content. An increased water/cement ratio will decrease strength and durability. Thus, in the future, it will be very important to use particle packing models to control the water demand and workability of ecological concrete. This can be done by using the cyclic design method. With this method, the amount of water in concrete mixtures can be reduced. The reduced amount of water and low water/cement ratio can be used to save a certain amount of cement when concrete is designed for a fixed strength class. In this way, the cement content in 3 concrete can be reduced from the prescribed minimum of 260 kg/m in ordinary concrete to 110 3 kg/m in ecological concrete. Experiments have shown that with this 57% reduction of Portland cement, material properties such as strength and modulus of elasticity can remain constant or even improve.

S.A.A.M. Fennis and J.C. Walraven

ACKNOWLEDGEMENTS This research was supported by the Dutch Technology Foundation STW, applied science division of NWO and the Technology Program of the Ministry of Economic Affairs. The experiments were performed by the technicians of the Stevin Laboratory of Delft University of Technology, which is very much appreciated.

REFERENCES [1] R.K. Dhir, M.J. McCarthy and K.A. Paine, Engineering property and structural design relationships for new and developing concretes, Materials and Structures 38 (275) (2005) 1-9. [2] F. de Larrard, Concrete mixture proportioning, A scientific approach, EF & Spon, London (1999). [3] W. Puntke, Wasseranspruch von feinen Kornhauf-werken, beton 52, H.5 (2002) 242-248. [4] I. Marquardt, Ein Misschungskonzept fr selbstverdichtenden Beton auf der Basis der Volumenkenngrssen und Wasseransprche der Ausgangsstoffe, Univ., Fak. fr Ingenieurswissenschaften, Fachbereich Bauingenieurswesen, Fachgebiet Baustoffe, Rostock, Germany (2002). [5] K.T. Miller, R.M. Melant and C.F. Zukoski, Comparison of the compressive yield response of aggregated suspensions: Pressure filtration, Centrifugation, and osmotic consolidation, Journal of the American Ceramic Society, Vol. 79, No. 10, (1996) 2545-2556. [6] H. Okamura, K. Ozawa, Mix design for self-compacting concrete, Concrete library of the JSCE, No 25, (Translation from Proc. Of ISCE, no 496/v-24, 1994.8), (1995) 107-120. [7] S. Mansoutre, P. Colombet and H. Van Damme, Water retention and granular rheological behavior of fresh C3S paste as a function of concentration. Cement and Concrete Research Vol. 29, (1999) 1441-1453. [8] M. Hunger and H.J.H. Brouwers, Development of Self-Compacting Eco-Concrete. In proceedings of the internationale Baustofftagung, Ibausil; 16., Bauhaus Universitt, Weimar, Germany. (2006) 1-1313 - 1-1320. [9] S. Palm and A. Wolter, Determining and optimizing the void filling of dry particle systems. Cement international, Vol. 7 (1/2009) 1-8. [10] G.H. Tattersall and P.F.G. Banfill, The Rheology of Fresh Concrete, Pitman, Boston (1983). [11] S.A.A.M. Fennis, Design of Ecological Concrete by Particle Packing Optimization, PhD Thesis, Delft: Delft University of Technology (2011). [12] F. Schwanda, Das rechnerische Verfahren zur Bestimmung des Hohlraumes und Zementleimanspruches von Zuschlgen und seine Bedeutung fr Spannbetonbau, Zement und Beton 37 (1966) 8-17. [13] W. Toufar, M. Born and E. Klose, Beitrag zur Optimierung der Packungsdichte Polydisperser krniger Systeme, Freiberger Forschungsheft A 558, VEB Deutscher Verlag fr Grundstoffindustrie (1976) 29-44. [14] J.D. Dewar, Computer Modelling of Concrete Mixtures, London: E & FN Spon (1999).