ASMS TECHNICAL NOTE NO 1

Concretes containing ground granulated iron blast-furnace slag as a Portland cement replacement exhibit properties not unlike those containing Portland cement alone.

With the exception of early age strength, most properties are enhanced by the replacement of cement with milled slag. This has resulted in slag blended cements replacing Type GP Portland cement in many applications and the production of special blends to produce properties in concrete not readily achievable with the available Portland cement types.

The data presented for slag cement blends is that available from commercial and research projects. The properties presented include strength, shrinkage, temperature rise and marine and sulphate resistance. Data on ternary blends is also presented.

Introduction

Ground granulated iron blast-furnace slag has been in use in Australia since the mid-sixties. On the east coast, slag as a cement alternative in the form of milled slag and in blended cements did not enjoy continuous availability until its reintroduction in 1982.

In Western Australia, the availability of slag cements has been uninterrupted since 1970 when milled slag was made available to concrete producers. Slag cement blends were introduced in 1974 and are currently available from two cement producers.

Currently, the use of granulated slag as a cement alternative is in high demand and approximately 250,000 tonnes of granulated slag are manufactured into blended cements by four cement producers.

Properties of Fresh Concrete Containing Slag

Workability

The lower relative density of ground slag causes an increase in paste volume. Compared to Portland cement, ground slag occupies about 9% more solid volume for the same mass. Thus, for equal weight of cement and equal water contents, this increase in paste volume generally benefits workability in mixes with low cement contents or where the aggregates lack the fine fraction.

Setting Time

The setting times for blended cements are generally longer than for Portland cements. As the percentage of slag increases in the blended cement or the concrete the setting or stiffening time extends. An increase in slag content from 35% to 65% increases initial set by approximately 60 minutes.

The delayed setting of blended cements allows concrete to be worked for longer periods. This is of benefit in avoiding formation of cold joints in large pours and in hot weather concreting.

Bleeding

The bleeding capacity and bleeding rate of concrete is affected by the ratio of the surface area of solids to the unit volume of water. When slag is used as a cement replacement, these effects depend on the fineness of the slag compared to the Portland cement and the combined effect of the total cementitious material.

If the slag is finer than the Portland cement, and substituted on an equal basis, bleeding is reduced. Conversely, if the slag is coarser, the rate of bleed increases. Blended cements containing slag and manufactured by the intergrinding process generally cause a reduction in bleeding tendency.

Properties of Hardened Concrete Containing Slag

Compressive Strength

It has been generally shown that concretes containing ground granulated blast-furnace slag as a cement replacement, at normal temperatures, develop strengths more slowly than that made from Portland cement [1-4].

The degree of decline in early age strength is a function of a number of variables. These include slag activity [5,6], method of proportioning and the slag content of the blend.

Slag blends manufactured by collective comminution, produce concretes with higher early strengths than concretes produced by blending the separate components. Figure 1 demonstrates the rate of strength development of the two types of blended cements. In each case the clinker component was identical and slag contents were 35% by mass.

Figure 1: Rate of strength gain

At early ages, the mode of manufacture of the blended cement can cause a difference of up to 10% in compressive strength for the same binder contents. This difference is of course dependent on the fineness of the cement and the slag as well as the reactivity of the two components.

Slag cements in current use produce concretes of equivalent 28 day compressive strengths for equal mass replacements of Portland cement. Slag contents are nominally at 25% by mass.

At normal curing temperatures, the slag component reacts at a slower rate than does Portland cement resulting in a lower compressive strength at 7 days. Figure 2 shows the difference in strengths for concretes containing binder contents between 300 and 450 kg/m3 and 80 mm slump.

Figure 2: Compressive Strength Performance

Higher replacement rates of slag for equal binder contents reduces strength at all ages as shown in Figure 3. This appears to be due to non-optimum gypsum contents particularly at replacement rates in excess of 50% [7].

Figure 3: Influence of Slag Content on Compressive Strength

Replacement rates above 40% by mass tend to produce concretes that exhibit low heat concrete behaviour. Data for a 30 MPa concrete mix design is presented in Table 1 and Figure 4. Air cooled slag was used as an aggregate.

Table 1: Effect of binder composition

Figure 4: Effect of binder

Flexural Strength

Blended cements containing slag generally yield higher moduli of rupture at ages beyond 7 days than do concretes without slag, particularly when slag is used at optimum proportions [6]. This is thought to be a result of the increased denseness of the paste in the concrete [8].

Drying Shrinkage

It is well documented in literature [9] that creep and shrinkage are related to he sulphate (gypsum) content of cement, and to cement chemistry and fineness.

The effect of slag content on drying shrinkage is shown in Figure 5. For 40 MPa concrete with slag contents varied between 0% and 80% by mass, the drying shrinkage:

1. increased by 11% at 14 days;

2. increased by 10% at 56 days; and

3. decreased by 7% at 365 days.

Figure 5: Influence of slag content on drying shrinkage

It follows that as slag replacement increases in the binder, there is a dilution of the sulphate content in the binder, causing higher volume changes to occur. The optimum sulphate level is that associated with the lower volume changes of concrete containing the cement alone.

Figure 6 demonstrates the effect of sulphate content on drying shrinkage for a binder containing 35% slag [7].

Figure 6: Influence of gypsum content on drying shrinkage

Creep

The influence of sulphate content on creep is shown in Figure 7. It can be seen that as the sulphate content increases, creep decreases. However, there does not appear to be an optimum sulphate content for creep as had been suggested by Alexander et al [9].

Figure 7: Effect of gypsum content on creep

Increasing slag content decreases creep. The effect of slag content in a 40 MPa concrete mix is shown in Figure 8.

Figure 8: Influence of slag content on creep

Thermal Cracking

Granulated slags and fly ashes have been commonly used as ingredients of blended cements as separate cementitious constituents to reduce the temperature rise in mass concrete.

It is important to note that, although the heat of solution method for determining the heat of hydration of cements indicates the total heat release potential of cement, it unfortunately does not indicate the rate of temperature rise, which is important in mass concrete applications.

Early age cracking occurs when the restrained strain during cooling exceeds the tensile strain capacity [10]. The restrained strain is the product of the coefficient of thermal expansion, the fall in temperature during cooling and the restraint factor [10].

Blended cements have a negligible effect on the coefficient of thermal expansion and the restraint factor, however, they do effect the temperature fall and reduce the tensile strain capacity of the concrete at early ages.

Ground granulated slag reduces the temperature rise and hence the temperature fall in comparison to Portland cement in concrete of equal strength grade. CIRIA Report 91 states that in large concrete pours the reduction in temperature fall is about 9% with 40% slag replacement and about 35% with 70% slag replacement [10]. The degree of temperature reduction is also dependent on the reactivities of the slag and of the cement used [6].

Comparison of temperature rise and temperature cycles in concrete under adiabatic conditions have demonstrated significant differences when blended cements are employed.

The effect of slag replacement is shown in Figure 9, 10 and 11 [13].

Figure 9: Temperature rise in concrete (slag, aggregate and ACSE cement)

Figure 10: Temperature / time curves under adiabatic conditions

Resistance to Sulphate Attack

Generally, slag cement concretes have been shown to have a higher level of sulphate resistance than Portland cement concrete. Frearson [14] found that even mortars containing 30% slag replacement were more resistant to sulphate attack than mortars made from Portland cement alone.

Figure 11: Comparison of calculated temperature cycles at centre of 500 mm slab for all cements

Comparative tests have shown that as slag replacement increases, slag cements acquire sulphate resistance similar to and in some cases exceeding that of sulphate resisting Portland cements. This is demonstrated in Figure 12.

Figure 12: Effect of slag content on expansion of mortar prisms (Type GP - slag blends)

The reason for the better performance of slag blends compared to Portland cement is unclear. It is thought that the performance is associated with the lowering of the C3A level in the concrete, reduction in the concentration of soluble calcium hydroxide in the paste matrix due to the reaction with the slag and changes to the internal pore structure of the paste [8].

It must be pointed out that the use of sulphate resisting cement or slag blends in concrete does not of itself confer immunity from sulphate attack. Low water/binder ratios and sufficient binder contents are essential for satisfactory performance. Vervbeck showed that water/cement ratio (cement content) is a significant factor [15].

Figure 13: Effect of cement type on the expansion of mortar prisms

Chloride Penetration

There is much evidence to support the belief that slag cements have higher resistance to diffusivity of chloride ions than do Portland cements [8, 16, 17, 18].

There is a suggestion that this is not due simply to reduction in impermeability and may be due to some form of chemical resistance or interaction with the entering chloride ions at the surface of the hydrating paste. This increase in resistance to chloride penetration improves as the slag replacement ratio increases.

Smolczyk [16] found that even at high water/binder ratios, high slag blends presented resistance to ingress of chlorides (see Figure 14).

Figure 14: Chloride-diffusion in concrete bars

Investigations [19] undertaken at the National Building Technology Centre in Sydney confirm overseas data for local materials. Lower levels of chloride penetration were obtained with three types of cement when slag replacement was greater than 60% by mass. Data is presented in Figure 15.

Figure 15: Effect of slag replacement

Durability

Carbonation

The rate of carbonation of slag cement concrete has been reported to be higher than that of Portland cement concrete, especially at high slag replacement values. In all cases, the rate of carbonation is related directly to the compressive strength of concrete [11].

Concretes stored in water or subjected to continuous moist curing, to reach their designed strength, generally show little or no carbonation [11]. In high quality concrete the penetration may only be 5 mm in 50 years of exposure [12].

This demonstrates the necessity for proper curing of all concrete. Mather found that cessation of curing at 3 days caused both slag cements and Portland cements to suffer strength loss to the same degree [20]. Thus for concrete when originally exposed, the carbonation depth is likely to be the same whether the binder contains slag blend or a Portland cement [16].

Sorptivity

Ho et al showed that concrete quality with response to interrupted curing is improved by the incorporation of slag in concrete [21].

The rate of water absorption by capillary action is significantly decreased when slag is used at replacement values of 35%. Sorptivity values for slag concretes were found to be similar to higher strength plain concretes. As the compressive strength of the concrete increases the influence of constituents reduces.

Properties of Ternary Blends

Ternary cements were first used in Australia in September 1966 by Specified Concrete Pty Ltd in Wollongong [22]. The cement binder was plant mixed and consisted of 40% Portland cement, 40% ground granulated slag and 20% fly ash. In March 1967, Australian Iron and Steel Pty Limited allowed the use of ternary cement in all concrete supplied to the steelworks.

In Sydney, the use of plant mixed ternary blends has been associated with the major concrete producers. The ternary binder is produced by adding fly ash to concrete containing slag cement blend. The quantity of fly ash incorporated depends on the mix philosophy employed and the grade of concrete being produced.

Typical binder contents and compressive strength results for commercial concrete are presented in Table 2. The data represents a production period of ten months.

Table 2: Composition and compressive strength of concrete containing ternary blend

Concrete Workability

Field results indicate that concrete containing ternary blends can be handled at lower slumps. Finishability is also reported to be better than for concrete where Portland cement alone is used.

Although both of these properties are affected by mix design, the increase in paste volume resulting from use of ternary blends, appears to improve workability and pumpability.

Setting Time

Blended cements generally cause setting times of concrete to increase. As the Portland cement content in the binder decreases, concrete setting time increases.

For ternary cement blends, this effect is similar to other blended cements. The increase in setting time being a function of binder composition. Typical values for concretes containing the same mass of binders, but of different composition, are presented in Table 3.

This effect on setting time generally decreases when suitable chemical admixtures are used in the concrete.

Table 3: Effect of binder composition on setting time for concrete containing 295 kg/m3 of binder

Strength Gain

For concretes containing equal masses of binder, the rate of strength gain for ternary blends is less than for Type GP cement at all early ages. This is shown in Figure 16. The rate of strength gain of ternary blends compared to that of Type GP Portland cement is demonstrated in Figure 17.

Figure 16: Rate of strength gain

Drying Shrinkage

There appears to be no significant difference in drying shrinkage between concretes containing ternary blends or those made with Type GP Portland cement.

The effect of binder composition on drying shrinkage is presented for 20 MPa and 40 MPa concretes in Figure 18 and 19.

Figure 17: Rate of strength gain compared to Type GP

Figure 18: Drying shrinkage (20 MPa concrete)

Figure 19: Drying shrinkage (40 MPa concrete)

Conclusions

The use of ground granulated blast-furnace slag in concrete allows the design of cements which provide the required concrete properties to suit the application required by the engineer.

To fully utilise slag cements, engineers and specifiers must acquire an understanding of the properties, potential and limitations of these materials.

Although slag cement concretes offer many attractive advantages, compared to Portland cement concretes, there are situations where they may not be the most appropriate materials.

References

1. Wainwright PJ and Tolloczko JJA, "Early and Later Age Properties of Temperature Cycled Slag-OPC Concretes", Second International Conference on Use of Fly Ash, Silica Fume, Slag and Natural Pozzolans in Concrete, Vol 2, Madrid, Spain, 1986.

2. Cook DJ, Hinczak I and Cao HT, "Development of Strength and Microstructure in BFS/OPC Blends", Concrete 87, Cement and Concrete Association, Brisbane, Australia, June 1987.

3. Reeves CM, "The Use of Ground Granulated Blast-furnace Slag to Produce Durable Concrete", How to Make Today's Concrete Durable for Tomorrow, Thomas Telford, London, May 1985.

4. Douglas E and Zebino R, "Characterisation of Granulated and Pelletized Blast-furnace Slag", Cement and Concrete Research, Vol 16, 1986.

5. Frearson JPH and Uren JM, "Investigations of Ground Granulated Blast-furnace Slag Containing Merwinitic Crystallisation", Second International Conference on Use of Fly Ash, Silica Fume, Slag and Natural Pozzolans in Concrete, Vol 2, Madrid, Spain, 1986.

6. Cook DJ Hinczak I. and Cao HT, "Heat of Hydration, Strength, and Morphological Development in Blast-furnace/Cement Blends", International Workshop on Granulated Blast-Furnace Slag in Concrete, Toronto, Canada, 1987.

7. Cook DJ, Hinczak I and Duggan R, "Volume Changes in Portland Blast-furnace Slag Cement Concrete", Second International Conference on Use of Fly Ash, Silica Fume, Slag and Natural Pozzolans in Concrete, Madrid, Spain, 1986.

8. Bakker RFM, "Permeability of Blended Cement Concretes", ACI SP 79-21, 1983.

9. Alexander KM, Wardlaw J and Ivanusec I, "The Influence of SO3 Content of Portland Cement on the Creep and Other Physical Properties of Concrete", Cement and Concrete Research, Vol 9, 1979.

10. Harrison TA, "Early-age Thermal Crack Control in Concrete", CIRIA Report 91, Construction Industry Research and Investigation Association London, 1981.

11. Osborne GJ, "Carbonation of Blast-furnace Slag Concretes", Durability of Building Materials, Vol 4, 1986.

12. Harrison TA and Spooner DC, "The Properties and Use of Concretes Made with Composite Cements", Interim Technical Note 10, Cement and Concrete Association, London, 1986.

13. Haywood Engineering Report, 1984.

14. Frearson JPH, "Sulphate Resistance of Combinations of Portland Cement and Ground Granulated Blast-furnace Slag", Second International Conference on Use of Fly Ash, Silica Fume, Slag and Natural Pozzolans in Concrete, Vol 2, Madrid, Spain, 1986.

15. Verbeck G, "Field and Laboratory Studies of the Sulphate Resistance of Concrete", Performance of Concrete - Resistance of Concrete to Sulphate and Other Environmental Conditions, University of Toronto Press, 1968.

16. Smolczyk HG, "Die Verwendung Von Hochofenzement Fur Stahlbeton Und Spannbeton", VI International Steelmaking Day, Paris, France, 1977.

17. Gjorv OE and Vennesland O, "Diffusion of Chloride Ions From Seawater Into Concrete", Cement and Concrete Research, Vol 9, 1979.

18. Roy DM and Idorn GM, "Hydration, Structure and Properties of Blast-furnace Slag Cements, Mortars and Concretes", ACI Journal, No 97, 1982.

19. Progress Report to Blue Circle Souther Cement Limited, "Marine and Sulphate Resistance of Blended Cements", National Building Technology Centre, Sydney, 1987.

20. Mather B, "Laboratory Test of Portland Blast-furnace Slag Cements", Journal of the American Concrete Institute, 1957.

21. Ho DWS, Hinczak I, Conroy JJ and Lewis RK, "Influence of Slag Cement on the Water Sorptivity of Concrete", Second International Conference on Use of Fly Ash, Silica Fume, Slag and Natural Pozzolans in Concrete, Vol 2, Madrid, Spain, 1986.

22. Private Correspondence - Mr J Visek, Consultant.

Ihor Hinczak

Please use your BACK button to return to the last page.

© Copyright 1996 QCL Group
ACN 009 658 520
Permission to display page on DPPEA site requested 3/6/98.