THE PERFORMANCE OF SUPPLEMENTARY CEMENTITIOUS MATERIALS (SCMs) AND WATERPROOFING ADMIXTURES F. Andrews-Phaedonos B.E. (Hons), M.I.E. Aust, C.P. Eng Technical Consultant-Concrete, GeoPave, VicRoads, Victoria

ABSTRACT Supplementary Cementitious Materials (SCMs), i.e., slag, fly ash and silica fume, are now recognised as an integral part of concrete technology and construction, due to their ability to enhance the properties of fresh and hardened concrete. Since their introduction in 1993 into the VicRoads concrete specification, a number of major structures have been successfully constructed utilising these cement replacement materials. An investigation was undertaken to evaluate their suitability under various curing regimes, including moist curing, polyethylene sheeting, curing compounds and wet/dry curing. The study has confirmed previous research results in that the use of SCMs in concrete results in significant reductions in the volume of permeable voids (VPV, as per AS 1012.21). The greatest VPV reductions occurred in concrete containing moderate single combinations of silica fume followed by fly ash and finally slag. Although triple blend mixtures can perform nearly as good as some of the single combination mixtures, four blend mixtures are not very efficient. SCMs appear to be more beneficial to the lower concrete grades than higher concrete grades, in regards to ultimate strength development, compared with conventional concrete. A hydrophobic pore-blocking admixture (HA) has been found to outperform both conventional and SCM concretes with far greater reductions in VPV. This admixture had little or no influence on strength development of concrete. A "waterproofing" admixture (WPA) has the least effect on reducing VPV compared with either SCMs or the hydrophobic pore-blocking admixture. Based on the efficiencies of the curing techniques with respect to strength development and VPV reduction, it has been recommended that polyethylene sheeting and curing compounds not be allowed for curing of concrete with high replacement of SCMs.

INTRODUCTION Supplementary Cementitious Materials (SCMs), i.e., slag, fly ash and silica fume, were introduced as cement replacement materials in the July 1993 version of the VicRoads structural concrete specification (Andrews-Phaedonos and Loh 1994), following an extensive literature review (Andrews-Phaedonos 1993). This review concluded that all the three materials impart favourable properties to both plastic and hardened concrete and further enhance the overall durability performance of structural concrete. The review further concluded that there are only a few disadvantages with these materials provided appropriate concrete designs are used and the usual techniques of placing, compacting and curing are followed. Since their introduction into the specification more than forty major bridge structures have been constructed utilising these cement replacement materials. In order to ensure the ongoing suitability of SCMs, an investigation was undertaken to evaluate the performance of concrete made with and without these materials, subjected to various curing regimes allowed by the standard specification. These included moist curing, polyethylene sheeting, curing compounds and wet/dry curing, resembling possible on-site curing conditions. The investigation formed part of a more comprehensive research study (4) which also evaluated the suitability of the simple test procedure AS 1012.21 as a quality control method for the determination of the volume of permeable voids (VPV). The performance of SCMs and conventional concrete was evaluated both in terms of VPV (measure of the interconnected void space) and the rate of strength development. The relative effectiveness of the various curing methods covered by the specification was also studied, including the sensitivity of particular SCM concretes to these curing methods. The relative attributes of a proprietary hydrophobic pore-blocking (HA) and a "waterproofing" admixture (WPA) were also investigated under the same curing and testing regimes. This paper presents the findings of this investigation. It should be noted that space restrictions prevent the presentation of additional data which is referred to in the overall discussion presented in this paper.

TESTS USED IN THIS STUDY Volume of Permeable Voids (VPV) - AS 1012.21 The method consists of drying the concrete specimens at 105o C ± 5 o C to constant mass, immersion of the specimens in water at 23 o C ± 2 o C, until constant mass is achieved (within 1 gram), then boiling the specimens for 5 hours and subsequently obtaining the mass in air and immersed mass in water. In this method, the "volume of permeable void space" is defined as that void volume which is emptied during the specified drying and filled with water during the subsequent immersion and boiling. Any voids which do not empty during drying, or which do not fill with water during wetting, would not be included in the interconnected pore space measured by this method. Capital investment in the simple equipment required for this method is very minimal and can easily be undertaken by any National Association of Testing Authorities (NATA) registered laboratory, with very little expertise for operation. A large number of specimens can be tested simultaneously and a full AS 1012.21 test would take approximately 7 to 10 days to complete.

Strength Development Compressive strength was measured at the end of various curing ages in order to provide a direct comparison between conventional and SCM modified concrete. The intention was also to obtain an

indication of the effectiveness of various curing regimes on strength development and the sensitivity of SCM - concretes to curing.

EXPERIMENTAL PROGRAM Concrete Mixtures and Grades In general, concrete mixtures were based on those supplied for VicRoads projects by commercial ready mixed concrete companies, which normally contain suitable water reducing and high range water-reducing admixtures. The program included concretes prepared from three different concrete grades with w/c ranging from 0.30 to 0.75. These were VR330/32, VR400/40 and VR470/55 representing concrete with 330 kg, 400 kg and 470 kg of cementitious material and 32 MPa, 40 MPa and 55 MPa standard cured strength (28 days moist-cured). Slag, fly ash and silica fume were used in single combination with Portland cement, as well as triple and four-blend combinations. In addition, a proprietary “waterproofing” admixture was used in single combination with Portland cement and in some SCM - concretes. A proprietary hydrophobic pore-blocking admixture was also used in single combination with Portland cement. Some fifty eight concrete mixtures, incorporating more than 1300 concrete cylinders and flat slabs, were made and tested in the test program. The overall program also included the extraction of some forty, 75 mm diameter x 150 mm long, concrete core samples from a number of newly constructed bridges (with and without SCMs), as well as extraction of concrete cores from previously investigated marine structures (South Gippsland Highway) ranging from 6 to 26 years old.

Materials The cement used complied with requirements for GP Portland cement given in AS 3972. Fly ash, slag and silica fume complied with AS 3582.1, AS 3582.2 and AS 3582.3, respectively. The fine aggregate was a natural sand and the coarse aggregate was a crushed basalt with a maximum particle size of 20mm. A proprietary water reducing admixture and superplasticicer were used in most concrete mixtures. The waterproofing admixture was in powder form and the hydrophobic pore- blocking admixture was supplied as a liquid which replaced the same amount of water. No air entraining admixture was used and the slump ranged between 35 mm and 200 mm.

Specimen Manufacture The concrete was mixed in accordance with AS 1012.2 in an open pan mixer. A number of cylinders (100 mm x 200 mm in size) were cast from each mixture. All cylinders were cast in two layers, each layer being compacted using a vibrating table. Two batches of cylinders prepared in the field were consolidated using hand-rodding. Concrete cylinder moulds were covered with mould caps and plastic sheeting for 18 to 24 hours after casting. They were then demoulded and transferred to their various curing regimes.

Curing Methods Following demoulding, concrete cylinders were placed into their respective curing regimes until required for testing, namely 3, 7, 28, 90, or 150 days. The following curing methods were utilised: a) b) c)

Moist curing in water maintained at 23oC ±2oC in accordance with AS 1012 (Curing method C1). Wrapped in polyethylene sheeting and stored out of doors for the whole duration of the required curing periods as indicated above (Curing method C2). Intermittent moist curing, to stimulate the most probable site conditions. Wet/dry conditions achieved by placement in water immediately after demoulding and exposing to the environment on alternate days (Curing method C3).

d)

Application of an approved curing compound immediately after demoulding followed by outdoor exposure for the duration of the required curing periods as stated above (Curing method C4).

Curing methods (C2, C3, C4) required concrete specimens to be mainly cured at an exposure site under uncontrolled conditions, with no deliberate attempt to control the temperature, thus intending to simulate environmental conditions that occur at typical construction sites. During the investigation period the average daily temperatures ranged between 6 oC and 27 oC, whereas the average relative humidity ranged from 51% to 70%. On very few occasions the lowest and highest temperature recorded were 1oC and 41 o C. The average monthly rainfall was 56 mm with the minimum and maximum values recorded at 12 mm and 197 mm respectively, during the same period. It is interesting to note that the average daily temperatures for the months of May to September ranged from 6 oC to 17 oC, whereas for October to April ranged from 10 oC to 27 oC. These certainly represent temperate environmental conditions. In addition daily minimum temperatures fell below 10 oC about 33% of the time during the study period whereas the daily maximum temperature exceeded 30 oC only about 12% of the time.

TEST PROCEDURES Compressive Strength The 100 mm x 200 mm concrete cylinders were capped with neoprene at the bottom end only and tested in compression in accordance with AS 1012.8. Testing was undertaken for the various curing regimes in accordance with the requirements of the VicRoads structural concrete specification at 3, 7 and 28 days as well as at 90 days. The 90 day strengths were to give a measure of the potential quality of the various concretes and to provide a measure of long-term strength development. Results reported are the average for two concrete cylinders. It should be kept in mind that cylinder strengths are only indicative of concrete strength in a real structure because of the difference in size.

Volume of Permeable Voids (VPV) As mentioned previously, the test method AS 1012.21 was used to determine the inter-related parameters of VPV and absorption. Testing was undertaken for the various curing regimes at 3, 7, 28 and 90 days for correlation with compressive strengths. Samples from concrete grade VR330/32, as well as some samples from other grades of concrete, were also tested at 150 days. This was to obtain a measure of VPV variation over those lengths of time. Each 100 mm x 200 mm concrete cylinder was cut into four equal 50 mm thick slices and the results were averaged.

RESULTS AND DISCUSSION Strength Development Strength development is a very important parameter in concrete construction, and normally has a great influence on the overall productivity on a construction site. In particular, it is of great importance in terms of form work and prop removal times without causing any damage to a structure, application of prestress for both pre-tensioned and post-tensioned precast units, lifting, and transporting of precast units, and of course possible early loading requirements superimposed on other cast in-situ concrete. Strength requirements are also very important in terms of satisfying design strength (as stated in drawings) and evaluating the long term performance of a particular concrete mixture, in satisfying minimum compressive strength specification requirements at 3,7 and 28 days for each concrete grade.

Effects of SCMs on Strength Development In order to obtain an accurate assessment of the relative effectiveness of the SCMs on the strength development of concrete, comparisons are made based on data generated from the specimens which were fully moist cured under laboratory controlled conditions (C1). Where necessary however, reference is made to data derived from other curing regimes.

Moderate replacement levels (single combinations with GP cement) The test results as illustrated in Figure 1 and Figure 2 confirm the well established view that slag, fly ash and silica fume have the capability to substantially improve the ultimate strength development of concrete (i.e., 28 days and beyond). Silica fume has proved to be the most effective, particularly at replacement levels of about 10 to 15% (see Figure 1). At the moderate replacement levels both fly ash and slag (i.e., fly ash up to 25% and slag up to 40%) were found to perform similarly, although fly ash appears to perform marginally better at its most efficient replacement level, which was found to be about 15 to 20%. The most efficient replacement level of slag was found to be about 30%. In relation to early age strength development (i.e., 3, 7 days), all slag and pozzolans at moderate replacement levels were found to be lagging behind conventional concretes (curing regime C1), with specimens subjected to curing regimes C2, C3 and C4 lagging further behind, although all of them still satisfy the early strength development (at 3, 7 and 28 days) as required by the specification (see Figure 5 and Figure 6). The only exceptions to this have been some compound-cured specimens (C4) which appeared to achieve the specified early age strengths one to two days later. It should be noted that both the fly ash and slag concretes generally catch up to conventional concrete strengths within the first 7 days, although at its most efficient levels (i.e., 15 to 20%), fly ash was found to overtake conventional concrete strengths within the first 5 to 7 days. The finer silica fume material was once again the best performer. This was found to catch up and overtake conventional concrete strengths within 3 to 7 days depending on the strength grade of concrete (see Figures 1 and 2). At the lower concrete grades (i.e., VR330/32), it was found that at replacement levels of 10 to 15% the early age strength development of silica fume concrete overtakes conventional concrete within 3 to 4 days. In fact, the overall test data indicate that all SCMs are more beneficial to the lower concrete grades than higher concrete grades, in regards to ultimate strength development compared to conventional concrete. The ultimate strength of slag and fly ash mixtures subjected to compound curing (C4), and to a lesser extent polyethylene sheeting (C2) was found to be slightly lower than similarly cured conventional concrete. Under these curing conditions, silica fume was found to suffer slightly higher ultimate strength reductions due to its greater susceptibility to loss of moisture (see Figure 1). This further highlights the importance of applying adequate early curing, particularly to slag and pozzolan concretes. Where silica fume is included in the concrete mixture as an additive (i.e., in addition to 100% GP cement), it appears to outperform both conventional and slag/pozzolans replacement mixtures. In fact a 10 to 15% addition of silica fume in a VR330/32 grade, would develop strengths comparable to specification requirements for the higher concrete grade VR470/55. The early age strength development would be very comparable or better than conventional concrete within hours.

High replacement levels At higher replacement levels the strength development of both fly ash and slag concrete (i.e., greater than 25% and 50% respectively) lags well behind both conventional, as well as moderate SCM - concrete mixtures (see Figure 2). Although ultimate strength and impermeability would be generally superior (i.e., beyond 90 days), the use of high replacement levels would result in considerable delays in formwork stripping times. This has to be considered very carefully when specifying such high replacement levels for cast in-situ works, and more importantly for precast concrete products including heat and steam curing. Although they begin to satisfy the specified strength developments at 28 days, they do not overtake

conventional concretes until after 50 to 90 days. In addition, at the higher replacement levels, slag performs better than fly ash (i.e., 65% and 40%, respectively). The cut off point at which both slag and fly ash fail to satisfy the specified early age strength development are at replacement levels of about 50% and 25%, respectively. The test data indicate that at the lower concrete grades (i.e., VR330/32 and VR 400/40, with higher replacement levels of both slag and fly ash), it may be possible to satisfy the early age strength development requirements if the w/c is reduced by about 0.05 to 0.10. With higher strength grades this would not be possible as the already low water content is needed to maintain the hydration. In general, the ultimate strength development of concrete specimens cured by curing compound (C4), polyethylene sheeting (C2) and wet/dry curing (C3) was found to be about 30%, 20% and 10% respectively, lower than the corresponding fully moist cured concrete (C1). In addition, compound curing (C4) and polyethylene sheeting (C2) result in a reduction in strength development of about 22% and 12%, respectively, when compared with wet/dry curing (C3). In terms of the specified strength development, compound cured specimens (C4) and some cured with polyethylene sheeting (C2), failed to satisfy the requirements. This clearly demonstrates the inadequate and unforgiving nature of curing compounds and the use of polyethylene sheeting alone (without wet hessian) towards high replacement levels of SCMs. All specified strength requirements, however, were met approximately 1 to 10 days later. As a result, it can be concluded that the use of curing compounds or polyethylene sheeting should not be allowed for curing concrete with high replacement levels of SCMs, particularly in aggressive environments (e.g., marine). Continuous moist curing should be the only acceptable curing regime.

Multiple blends Triple blend concrete mixtures at moderate replacement levels satisfy the specification requirements for strength development (see Figure 3). However, as with single replacement combinations, they lag behind conventional concretes at the early ages, although they catch up at about 7 days. The strength development of triple blends is comparable to moderate replacement slag concretes and marginally lower than the corresponding fly ash concretes. This is mainly influenced by the slightly higher total replacement levels associated with the more efficient triple blend mixtures compared to the single combination concrete mixtures. The most efficient combination of triple blends at the moderate replacement levels, with regards to strength development appear to be the combinations 75% GP cement/17% fly ash/8% silica fume and 68% GP cement/24%slag/8% fly ash. Triple blends with higher replacement levels generally behave in a similar manner to single combination, high replacement level mixtures, with lower early strengths (ie. 3,7 days), although they pick up and overtake conventional concrete after about 50 to 90 days. The most efficient combination of high replacement triple blend appears to be 30% GP cement/60% slag/10% silica fume. A single four blend mixture consisting of a combination of slag/pozzolans with a replacement of up to 50% (i.e., 50%GP cement/25% slag/17% fly ash/8% silica fume) was also tested. However, this also appears to behave in the same manner as other high replacement concrete mixtures. Thus it can be concluded that multiblend mixtures should be subject to the same curing requirements and limitations which apply to both moderate and high replacement concrete mixtures, as discussed previously.

"Waterproofing" admixture (WPA) Examination of Figure 4 reveals that the incorporation of WPA into conventional concrete, results in a noticeable increase in strength, particularly in the early ages, as a result of its ability to enhance the hydration process. The early strength development is at least 25% higher when WPA is used, whereas the strength increase at later ages is of the order of about 12%. It was found that the optimum dosage of this particular WPA is about 1% by mass of cement (manufacturer's recommendation 1.5%) which may result in even higher strengths particularly for lower concrete grades. The test data also reveals that the addition of WPA to conventional concrete, increases the early age strengths by a similar margin for all grades of concrete. With regards to ultimate strengths however, it was found that the addition of WPA has a greater effect on the lower concrete grades, with a diminishing effect as the concrete grade increases. This is related to the w/c and the availability of adequate moisture to maintain the hydration reaction, both in the early stages, as well as at later stages. In contrast to the higher concrete grades, VR330/32 has sufficient

water to maintain adequate hydration at all stages of strength development. It was also found that the incorporation of WPA into SCM concretes does not improve strengths in the same way as with conventional concrete. In the case of moderate replacement SCM concrete the effect is less profound with marginal increases of about one third those associated with conventional concrete. Once again the effect on ultimate strength diminishes with increased concrete grade. With higher replacement levels of SCMs, the addition of WPA results in marginal reductions in strengths at all ages. This is related to the fact that both SCMs and the WPA are competing for the same product of cement hydration, namely, Ca(OH)2 and water which is of course badly needed by the SCMs particularly at the higher replacement levels. The quicker WPA reaction diminishes the effect of the slower reacting SCMs. It is quite evident from the above discussion that WPA can be very beneficial for strength development, when added to conventional concrete and less so for SCMs concrete. It is therefore considered that further work is required with a variety of WPAs in combination with SCMs, before embarking on any such use in practice. Nevertheless, the above discussion and test results provide a good initial indication as to their possible performance.

Hydrophobic admixture (HA) The proprietary hydrophobic admixture was found to have little or no adverse effect on compressive strength. The strength development of hydrophobic admixture concrete satisfies the specification and was found to be similar to conventional concrete, particularly for the lower concrete grades. For the higher concrete grade it resulted in lower strength than conventional concrete. However, these differences are not considered to be of any practical significance as they still satisfy the specification requirements. By contrast, the reductions achieved in apparent VPV by hydrophobic concrete are very substantial compared with conventional, as well as SCM concrete. This is discussed further under the effects of hydrophobic admixtures on VPV.

Volume of Permeable Voids (VPV) The importance of using the VPV as an indicator of permeability and potential long-term durability of concrete, is demonstrated by its significant correlation with the percentage of water absorption of the concrete. The VPV essentially indicates the interconnected space within the concrete which accommodates the absorbed water and moisture movement in the concrete. The significance of determining the VPV is further highlighted by its direct relationship with the percentage of total volume of hardened concrete represented by water voids and other possible interconnected voids or passages which act as a path for future moisture movement. Factors which may influence the total interconnected porosity include effects of poor mixture proportions, high w/c, poor curing, shrinkage, which results in interconnected microcracking (more so in lower grades concrete), poor compaction, and porous aggregates.

Effects of SCMs on VPV The test data on VPV presented in Figure 5, Figure 6 and Figure 7 confirm the growing body of test results, from previous research such as on water and chloride permeability and porosimetry testing, that the use of SCMs in concrete results in significant reductions in permeability of aggressive agents. The lower penetrability of SCMs concrete compared to plain concrete is of course attributed to the increased denseness, due to the significant refinement of the pore system, both in the bulk paste as well as, the transition zone between the cement paste and aggregates. Depending on the type, replacement level and combinations of SCMs used, reductions in VPV of up to 35% have been recorded in this investigation. The greatest VPV reductions occurred in concrete containing silica fume, followed by fly ash and finally slag. As mentioned earlier, in terms of the bulk nature of VPV of concrete such reductions in penetrability are quite substantial.

Effects of silica fume Silica fume mixtures in single combination with GP cement have shown the greatest reductions in VPV, thereby confirming the superiority of silica fume in both strength development and lower VPV over the other SCMs. The results presented in Figure 5 indicate an average reduction in VPV of about 20%, with some silica fume mixtures decreasing VPV by as much as 35% (particularly the higher concrete grades). This substantial reduction is very similar for both the 10% and 15% replacement levels, although the higher replacement level is marginally ahead. The use of silica fume as an additive (i.e. 10% to 15% in addition to 100% GP cement) results in slightly higher reductions in VPV compared to the replacement mixtures.

Effects of fly ash Fly ash mixtures in single combination with GP cement have exhibited substantial reductions in VPV. As shown by Figure 6 the average reduction in VPV after 90 to 150 days of moist curing is of the order of 15% compared to plain concrete mixtures. The most efficient fly ash replacement level of about 17% by mass of cement, exhibited VPV reductions well in excess of 25% compared to plain concrete mixtures.

Effects of ground blast-furnace slag As indicated in Figure 7 the incorporation of slag in single combination with GP cement has resulted in average reductions in VPV of about 10% compared to plain concrete and as much as 20% in some cases. Although this is not as dramatic as the reductions achieved by the use of silica fume or fly ash, in the context of the parameter being measured, these are considered quite significant. It is also interesting to note that slag in combination with the lower end of w/c have resulted in VPV reductions well in excess of 30%.

Effects of multiple blend Triple blends are generally characterised by reduction in VPV of the order of 10-15% compared to conventional concrete mixtures. However, some of the more efficient triple blend mixtures such as 30% GP Cement/ 60% slag/ 10% silica fume or 75% GP Cement/ 17% fly ash/ 8% silica fume exhibit slightly higher reductions. Four blend mixtures are not very efficient. Essentially they are performing similar to conventional concrete with only marginal decrease in VPV of the order of about 3%.

High replacement levels Generally mixtures with high replacement levels of SCMs have exhibited reductions in VPV of the order of 10% compared to conventional concrete. Of course high replacement triple blends such as 30% GP cement/ 60% slag/ 10% silica fume exhibit slightly higher reductions.

Effect of field-made concrete cylinders As expected field made cylinder specimens (compacted by rodding and cured in accordance with AS 1012.8) were found to exhibit higher VPV values compared to laboratory controlled specimens with the same w/c. The mixtures used were 90%GP cement/10% silica fume (w/c of 0.36) and 66% GP cement/22.5% slag/11.5% fly ash (w/c of 0.50), but restrictions on space for this paper prevent the presentation of more data.

Effects of hydrophobic admixture (HA) Examination of test results presented in Figure 8 reveals that the reductions in apparent VPV achieved by hydrophobic concrete under similar testing conditions are very substantial compared to both conventional

and SCMs concrete. In fact the average reduction in apparent VPV of about 30% achieved by the hydrophobic pore-blocking admixture is even larger than the average reductions offered by silica fume concrete. Some individual hydrophobic concretes have even reduced apparent VPV by as much as 38% compared to conventional concrete. By contrast, as discussed previously this proprietary hydrophobic admixture has little or no influence on strength development of concrete.

Effects of “waterproofing” admixture (WPA) Test results indicate that WPA has the least effect on reducing VPV compared with either SCMs or the hydrophobic pore-blocking admixture. Examination of Figure 9 reveals that the effect of WPA on VPV is very similar to its effect on strength development. As with strength development, it has its greatest effect on the lower concrete grade (i.e., VR330/32) with reductions in VPV of the order of 10%. However, for the higher grades of concrete its effect diminishes even further with reductions in VPV limited to about 5%. These results are clearly influenced by the amount of moisture available to maintain the hydration process and allow further refinement of the pore system. The limited effect of WPA compared with SCMs and hydrophobic pore-blocking admixture is mainly related to the coarser microstructure developed as a result of the greater initial heat generation and therefore its rapid hydration process. The addition of WPA in moderate replacement levels of SCM concrete has only a minimal effect on VPV reductions over and above the individual effects of the SCMs. Additional reductions in VPV were only limited to about 5%. In the case of higher replacement levels of SCMs, the addition of WPA has a very negative effect, with an increase in VPV of the order of 15%. This is closely related to the amount of moisture available and the fact that both SCMs and the WPA are competing for the same hydration product, namely, Ca(OH)2. The quicker WPA reaction diminishes the effect of the slower reacting SCMs. Although WPA may be a useful additive for use in conventional concrete, when a higher early strength is required, particularly at lower concrete grades, it is not as effective in delivering significant reductions in VPV due to the resulting coarser microstructure. In addition, its effectiveness with SCM concrete and in particular at higher replacement levels appears to be limited.

CONCLUSIONS All SCM materials at moderate replacement levels satisfy the early strength development as required by the VicRoads specifications. SCMs appear to be more beneficial to the lower concrete grades than higher concrete grades in regards to ultimate strength development compared to conventional concrete. At the higher replacement levels, the early strength development of concretes with fly ash and slag and triple blends lags behind both conventional and moderate SCMs concrete mixtures. The strength development performance of moderate triple blends is comparable to moderate replacement levels of slag and marginally lower than the corresponding fly ash concretes. The use of SCMs in concrete results in significant reductions in permeability to aggressive agents. The greatest VPV reductions occurred in concrete containing moderate single combinations of silica fume followed by fly ash and finally slag. Moderate triple blend mixtures can perform nearly as well as some of the single combination mixtures. Quaternary blend mixtures are not very efficient. Mixtures with higher replacement levels of SCMs have exhibited reductions in VPV similar to those of moderate replacement by fly ash or slag. The hydrophobic pore-blocking admixture has been found to outperform both conventional and SCM concretes with far greater reductions in apparent VPV. By contrast this admixture has little or no influence on strength development of concrete.

The "waterproofing" admixture (WPA) has the least effect on reducing VPV compared with either SCMs or the hydrophobic pore-blocking admixture. Although it may be a useful additive for use in conventional concrete, when a higher early strength is required particularly at lower grades, it is not effective in delivering significant reductions in VPV due to the resulting coarser microstructures. In addition, its compatibility with SCMs concrete and in particular at higher replacement levels appears to be very limited. The sensitivity of moderate replacement SCMs to various curing methods is similar to conventional concrete. However, at higher replacement levels, SCMs are extremely sensitive particularly to less efficient curing methods such as polyethylene sheeting and curing compounds. Curing compounds and polyethylene sheeting should not be allowed for use on high replacement SCM concrete.

REFERENCES ANDREWS-PHAEDONOS, F. AND LOH, H. H.(1994), "Durability Provisions in VicRoads Structural Concrete", AUSTROADS, Bridges Conf., Melbourne. ANDREWS-PHAEDONOS, F.(1993), "Research Proposal, Performance of Concrete Containing Supplementary Cementitious Materials", VicRoads Internal Report, VicRoads. ANDREWS-PHAEDONOS, F.(1993), " Use of Supplementary Cementitious Materials in ConcreteGGBF Slag, Fly Ash, Silica Fume", Internal Report, VicRoads, Melbourne. ANDREWS-PHAEDONOS, F.(1996), "Establishing the Durability Performance of Structural Concrete", Internal Report, VicRoads, Melbourne, 1996. MALHOTRA, V. M. (Editor)., "Fly Ash, Silica Fume, Slag and Natural Pozzolans in Concrete", Proc. Istanbul Conf., ACI SP-132, Vol. I and II. MALHOTRA, V. M. (Editor).(1989), " Fly Ash, Silica Fume, Slag and Natural Pozzolans in Concrete", Proc. 3rd Int'l Conf., Trondheim, Norway. MALHOTRA, V. M. (Editor). (1994), " Durability of Concrete", 3rd Int'l Durability Conf., Nice, France, SCANLON, J. M. et al.(1994), "Use of Fly Ash to improve Concrete Durability By Reducing Permeability".

AUTHOR BIOGRAPHY Fred Andrews-Phaedonos graduated from Monash University in 1978 with an honours degree in Civil Engineering. Since then he has worked for VicRoads, mainly in bridge and concrete related areas. At present he is a member of the GeoPave Department of VicRoads, where he is a technical specialist in the areas of concrete technology, concrete durability, diagnostic assessment, protection and repair of concrete. He is a member of several Standards Australia and AUSTROADS technical committees and was a member of the organising committees for the RILEM and AAR international conferences held in Melbourne in 1992 and 1996 respectively. He is an elected member of the Federal Council and current President of the Victoria Branch of the Concrete Institute of Australia. He is the author or co-author of more than 35 technical papers presented at both national and international conferences in Australia and overseas, and numerous major technical reports released within VicRoads and elsewhere.

ACKNOWLEDGEMENT The author wishes to thank the Chief Executive, VicRoads for permission to publish this paper. The views of this paper are those of the author and do not necessarily represent those of VicRoads.