Concrete performance using BFS cement

Eberhard Lang presents experimental data on high-performance concrete using blastfurnace slag cement. Durability, low heat of hydration and workability are notable features: superplasticisers and silica fume can be used to further modify concrete properties.

Results of investigations in laboratory scale as well as long experiences in practice indicate the special benefits of blastfurnace slag cements. These are, besides others, their comparatively low capillary porosity, their high resistance against sulphate attack, seawater or other aggressive attack, against alkali-aggregate reaction and to the diffusion of chlorides into the concrete. Recent research has also shown a high binding capacity of chlorides in blastfurnace slag cement pastes. A further aspect of importance for the corrosion protection of reinforcement is the comparatively low electrical conductivity of concrete made with blastfurnace slag cement. Due to their low heat of hydration, blastfurnace slag cements help to prevent cracks in concrete structures resulting from temperature stresses at early ages.^{1, 2}

This paper described researches to use some of these special properties in high performance concrete and shows some practical experiences.

Experimental Details

Materials
Blastfurnace slag cement with different slag content (48%, 63% and 78%) and two different strength classes (32,5 and 42,5) were used. Table 1 shows the chemical composition and Table 2 the most important physical properties. Silica fume was used in form of a slurry with a water content of 50%. The coarse and fine aggregates consisted of natural gravel and sand from the Rhine area and crushed air-cooled blastfurnace slag (bfs) with a particle size distribution of 2/8mm and 8/16mm. As admixtures a sulphonated melamine formaldehyde-based or a Naphthalensulphonic acid based superplasticiser and in some cases a retarder have been used.

Table 1: Checmical composition of the blastfurnace slag cements
* NW - low hear, HS - high surface resistance, NA - low effective alkali content

Table 2: Physical properties of blast furnace slag cements

Mix Design
A selection of mix design used for the durability tests of some high performance concretes are summarised in Table 3.

Table 3: Mix design of selected high performance concretes

Test Methods
The fresh and hardened concrete properties were tested in accordance with the German standard DIN 1048. The dynamic modulus of elasticity was tested with a Grindo-Sonic MK 4x-instrument.

The pore size distribution was measured by a Mercury pressure porosimeter series 4000 by Carlo Erba in the range 1.85 to 7500nm radius. The results obtained are cumulative pore size distribution and the total porosity.

The freeze-thaw and deicing salt resistance of concrete was determined according to prEN 12390-9 - CDF-test (Capillary suction of Deicing solution and Freeze thaw test).

The equipment used for measuring the penetration of organic liquid is shown in Figure 1. ³

The samples were drill cores with heights of 150mm and diameters of 80mm which were obtained at the 7-day moist storage. The cores were stored for 56 days at 20°C and 65% relative humidity until tested. In order to achieve a one-dimensional transport process the cores were sealed at their circumferential surfaces. For this purpose they were coated with epoxy resin adhesive and at the same time wrapped in special steel foil. This surrounded also a metal cylinder served to hold a burette with graduations and also made it possible to distribute the test liquid uniformly over the sample surface.

After sealing, the core and metal cylinder were also wrapped with rubber tape to ensure that the fresh and the hardened epoxy resin were constantly pressed against the circumferential surface of the core. This ensured that the circumferential surface was impermeable even when the resin was dissolved by the test liquid.

During the test the burette was filled rapidly with the test liquid to a level of 1400mm above the sample, and the exact initial level recorded from the graduations. During the subsequent 72-hour test the quantity of liquid which had penetrated could be read at any time from the graduations, which made it possible to follow the course of the penetration. After completion of the test period of 72 hours the cores were unwrapped and split parallel to the longitudinal axis, and the visible penetration front of the liquid was recorded.

Results and discussion

Selection of superplasticiser
In laboratory scale different superplasticisers were tested in combination with the used blastfurnace cements (Figure 2).

In all cases the maximum of the allowed plasticiser content (according the data sheet of the producer) and 50% of this content were tested. The influence of the tested plasticiser on the change of the flow point is very different and shows a very wide range. On the other hand the influence of the kind of blastfurnace cement is relatively small.

Optimisation of superplasticiser and silica fume content
A statistical work schedule allowed an optimisation of the kind and the content of superplasticiser and the content of silica fume. Figure 4 shows as an example the result for one type of plasticiser and the development of the 7 day strength.

Properties of fresh concrete
The consistence of the fresh concretes is shown in Table 4. All concretes are classified according to the flow classes F3 and F4. The decrease of the cement content from 455kg/m³ to 435kg/m³ reduces the flow class for one grade, but the decrease to 415kg/m³ gives nearly the same result as with 435kg/m³. The consistence of the concrete 1 with the retarder belongs after 45 minutes to the same flow class as after 10 minutes. Without retarder the consistence class decreases for one class in this period. The consistence of the concrete 2 with the absorbent air cooled blastfurnace slag decreases for one class, too.

Table 4: Consistence and air void content of the fresh content
* Consistence a₁₀ and a₄₅: time in minutes after mixing

Strength, modulus of elasticity
The development of strength is given in Figure 5, above. The results show the good long-term strength development of concretes made with blastfurnace cements.

The guideline for high strength concrete in Germany includes the strength classes from B 65 to B 115. In accordance with this guideline the criteria for compressive strength for initial testing depend on the strength class. For strength class B 75 the criteria is 87MPa and for B 85 it is 99MPa. These results show that it is possible to produce a high strength concrete with all types of blastfurnace slag cements, even with slag contents >70%. Only the 1 day strength of this cement type is lower.

The dynamical modulus of elasticity increases with the compressive strength, Table 5.

Table 5: Relation between compressive strength and modulus of elasticity

Porosity
Figure 6 gives an overview about the pore size distribution of a standard mortar as given in Table 2 (w/c-ratio 0.50) and three mortars without, with 4% and with 10% silica fume. The silica fume was used as a replacement for cement. These three mortars had the same consistency with the addition of superplasticiser. The best workability is given with a mortar with 4% silica fume which also has the lowest w/c-ratio and the lowest volume of capillary pores. The pore size distribution of blastfurnace slag cement mortars with and without silica fume is very similar. The porosity in the hardened cement paste of the concretes is comparable with the results in Figure 6.

As shown in Figure 7, the volume of capillary pores (30-100nm) one day after producing of the concrete No. 7 is high (11.7%). Already after 7 days the volume of capillary pores is very low, but in the range from 10-30 nm the volume has increased. After 28 days the porosity has increased in this range of 10-30 nm. Between 28 and 180 days the change of porosity is insignificant.

Shrinkage
The drying shrinkage of concrete depends upon the unit amount of cement and aggregate, the w/c-ratio, the types and fineness of cement, the types and unit amount of admixture and other. Figure 8 shows the shrinkage of the concretes No. 1 - 4.

The drying shrinkage of concrete prepared by demoulding after 1 day, curing in water for 6 days, and drying in the atmosphere of RH of 65% at 20°C decreases with the decrease of the cement content. In comparison with the results of Mori the concretes with blastfurnace slag cement show a lower drying shrinkage than the Portland cement at the same w/c-ratio.

Carbonation
In general in the atmosphere of RH of 65% at 20°C blastfurnace slag cements with high slag content have a higher carbonation depth than Portland cements. In the natural atmosphere this difference is much lower. In comparison with ordinary concrete the carbonation depth of HPC is very low, as shown in Table 6. The curing conditions are the same as for the measuring of shrinkage with a normal CO₂-content of 0.03%.

Table 6: Carbonation depth, curing in natural atmosphere

Freeze-thaw and deicing agent resistance
Concrete made with high slag blastfurnace cements in natural atmosphere leads to lower carbonation depth than in laboratory conditions. A comparable tendency is well known for the freeze-thaw-resistance. Concretes with high-slag blastfurnace cements have proved their durability by lasting for decades under high working stress, for example in sewage treatment plants or in sea locks. Even without artificial air voids, these concretes have proved a good resistance to freezing and thawing cycles, to seawater or deicing salt attack.

The results of measurement of the freeze-thaw and deicing agent resistance with a solution of water and 3% by weight of NaCl is shown in Figure 9.

All the concretes were prepared without an AE agent and the HPC without silica fume. The resistance to freezing and thawing is significantly improved by lowering the w/c-ratio. The resistance of the HPCs with silica fume is improved additionally. A sufficient resistance to freezing and thawing is given in practice already for concretes with blastfurnace slag cement and w/c-ratio of 0.50. The decrease of the w/c-ratio to the range of 0.3 improved the resistance significantly.

In laboratory tests ordinary concretes with high slag blastfurnace cements lead to a higher scaling in the first freeze-thaw cycles than Portland cement. This scaling ranges between 0.1 to 0.3mm cement grout from the surface. Afterwards the scaling rate decreases significantly. HPC with high slag blastfurnace cement does not show this effect. With these strong laboratory conditions the scaling rate is uniform and the freeze-thaw-resistance very high. The scaling of concrete containing a cement with 50% blastfurnace slag is very low.

Impermeability to organic liquids
In the chemical industry catchment tanks made of uncoated concrete are secondary barriers for limited periods of time (in general 72 hours) if the storage tank has a damage. The capillary porosity of concrete with blastfurnace slag cement is lower than of concrete with Portland cement and the chemical resistance in general is higher. Therefore a lot of catchment tanks in Germany are carried out with blastfurnace slag cement. For this application the impermeability of HPC was tested. The research in this field is going on. Figure 10 gives initial information on the penetration depth of four different chemical liquids. The ordinary concrete has a cement content of 320kg/m³ and a water/cement-ratio of 0.50.

Application of blastfurnace cement in practice
Table 7 show a mix design and some results of the production control for the use of blastfurnace cement with high sulphate resistance (HS) and a low alkali content (NA) for the production of pipes for drinking water.

Figure 11 shows examples of pipes for drinking water produced with the mix design according to Table 7 (figures by E+F GmbH Rohrwerk Epiton). The pipes show a very high durability and the are well introduced in the market.

Conclusions

Based on the results presented in this paper it can be stated that High Performance Concrete can be produced using all types of high slag blastfurnace cement (CEM III/A 42.5, CEM III/A 32.5, CEM III/B 32.5). These concretes have a good workability and the well known special properties of durability are additionaly improved.

The tested superplasticiser shows a very different efficiency depending on the kind of clinker, the fineness and the slag content.

The use of silica fume in combination with blastfurnace cement is only necessary for high early strength.

The low heat of hydration is an important advantage to minimise the risk of microcracks and thermal shrinkage. Porous air-cooled blastfurnace slag as aggregate can be used for an 'internal curing'. Practical experiences confirm the laboratory test results.

Porous air-cooled blastfurnace slag as aggregate can be used for an 'internal curing' of HPC.

The freeze-thaw resistance and the carbonation rate, sometimes a problem for high-slag blastfurnace cement in laboratory tests, is improved significantly.

Literature

1. Geiseler, J., Kollo, H., and E. Lang: 'Influence of Blast Furnace Cements on Durability of Concrete Structures' ACI Materials Journal, 92 (1995) 3, (May-June), S. 252/257

2. Lang, E.: Blastfurnace Cements in 'Structure and Performance of Cements,' 2nd Edition by J. Bensted and P. Barnes, Publishing house: Spoon Press, London and New York, 2002

3. DAfStb-Richtlinie 'Betonbau im Umgang mit wassergefährdenden Stoffen.' Beuth-Verlag Berlin 2004