SCALING RESISTANCE OF CONCRETE CONTAINING SLAG CEMENT: A CRITICAL REVIEW

Transcription

1 SCALING RESISTANCE OF CONCRETE CONTAINING SLAG CEMENT: A CRITICAL REVIEW Fatih Bektas Assistant Professor Texas State University 4214 Roy F. Mitte San Marcos, TX Phone: fbektas@txstate.edu Peter C. Taylor Associate Director National Concrete Pavement Technology Center Iowa State University 2711 South Loop Drive, Suite 4700 Ames, IA Phone: ptaylor@iastate.edu Kejin Wang Associate Professor Iowa State University 492 Town Engineering Ames, IA Phone: kejinw@iastate.edu Word Count: 3,492 text words plus 1,000 for figures/tables (4 250) = 4,492 total Submission date: July 27, 2009 Revised: November 13, 2009 Corresponding author

2 ABSTRACT Concrete containing slag cement (i.e., interground with clinker or replacing ordinary portland cement) has been widely used for many years and has provided proven long-term performance. On the other hand, there are concerns about the deicer salt scaling resistance of concrete containing high amounts of slag (e.g., 50% and up). These concerns are generally based on laboratory results obtained from ASTM C672 procedure. However, laboratory results are not generally consistent with field observations. A literature survey has been conducted to compile laboratory data from reported ASTM C672 tests along with field performance data. The objective was to seek evidence that would support the selection of an upper limit for slag dosage in concrete that will be exposed to deicer salts. The findings were that when using up to 50% slag it is possible to produce concrete that is resistant to salt scaling. However, slag content is not the sole factor controlling scaling resistance but construction related issues such as finishing and curing are also significant. In addition, the current ASTM C672 procedure reportedly does not reliably predict field performance of concrete containing slag.

3 INTRODUCTION Ground granulated blast furnace slag, known as slag cement or slag, is widely used in portland cement concrete and can be used at dosages up to 75% by mass of cementitious material. Concrete containing slag has a proven history of long-term performance in many environments and applications. With the growing emphasis on sustainable infrastructure development, slag is increasingly attractive to the construction industry, agencies and the public. However, the performance of concrete containing slag is still debated in one area salt scaling resistance. In colder regions of the North America various salts are extensively used as deicing agents on pavements and bridge decks. Damage due to salt scaling is a common problem in portland cement concrete and there is a perception that this is exacerbated when slag is used in the mix, particularly at higher percentages. The Federal Highway Administration (FHWA) limits slag use to 25% for concretes that will be exposed to deicer salts as opposed to 50% for areas not exposed to deicer salts (FHWA 1999). Similarly, the Ontario Ministry of Transport (MTO) allows a maximum of 25% slag replacement for bridge decks and barriers (Boyd and Hooton 2007). Concrete containing slag is claimed to perform poorly in salt scaling tests when compared to ordinary portland cement concrete (Marchand et al. 1994; Valenza and Scherer 2007). A review by Taylor et al. (2009a) reported that the relatively poor performance of systems containing slag was potentially due to a combination of: changes in surface microstructure; changes in hydration and bleeding rates coupled with different finishing and curing practices; accelerated carbonation; and changes in pore solution which influence osmosis. The reported poor performance is usually based on laboratory testing using ASTM C672 Standard Test Method for Scaling Resistance of Concrete Surfaces Exposed to Deicing. However, the ASTM C672 exposure regime is reported to be harsh and does not simulate field conditions, and is likely to reject acceptable systems (Bouzoubaa 2008). Reports have shown that the field performance of concrete containing more than 25% slag has been adequate (Marchand et al. 1994; Luther et al. 1994; Boyd and Hooton 2007). In addition, some researchers have reported concrete containing high slag contents (i.e., 50%) can also exhibit sufficient salt scaling resistance, even in the ASTM C672 testing environment (Luther et. al. 1994). This indicates that the testing procedure has the potential to provide misleading data. A new test method is therefore required that produces findings more consistent with field observations. In addition, a review of the data is needed to assess whether there is a logical basis on which to limit the maximum slag dosage that should be applied for concrete mixtures exposed to deicing salts. This paper addresses the latter need. SALT SCALING RESISTANCE TESTS The analysis in this paper uses data reported from tests conducted in accordance with ASTM C672 and the Ontario equivalent, MTO LS-412. ASTM requires the specimens have a surface area of at least 0.045m 2 and a thickness of at least 75mm. The MTO LS-412 method specifies a specimen size of mm 3, or an equivalent surface area. In both methods the specimens are demolded after 24h and subjected to standard moist curing for 13 days. Thereafter, the specimens are stored in air for 14 days at 23.0±2.0 o C and % relative humidity. This limited curing time is therefore likely to be biased 1

4 against mixtures containing supplementary cementitious materials such as slag because of their slower hydration rates. At 28 days old the concrete under evaluation is then covered with a 6mm deep salt solution. The ASTM procedure recommends a 4% CaCl 2 solution while the MTO LS-412 uses a 3% NaCl solution. The specimens are subjected to 50 freeze-thaw cycles, each cycle lasting 24 hours. Each cycle comprises 16 to 18h of freezing at around -18 o C and 6 to 8h of thawing at standard room temperature of 23 o C and 45-55% relative humidity. Every five cycles the salt solution is washed off and replaced, and a surface assessment is done in accordance with the standard. MTO LS-412 also assesses the damage based on the dry mass of material lost from the surface. A loss in excess of 0.8 kg/m 2 after 50 freeze-thaw cycles is considered as failure. On the other hand, ASTM C672 uses a visual evaluation on scale from 0 to 5 (Table 1). TABLE 1 ASTM C672 Visual Assessment Rating Condition of Surface 0 No scaling 1 Very slight scaling (3 mm depth, max, no coarse aggregate visible) 2 Slight to moderate scaling 3 Moderate scaling (some coarse aggregate visible) 4 Moderate to severe scaling 5 Severe scaling (coarse aggregate visible over entire surface) GENERAL TREND OF THE LABORATORY TEST RESULTS Concrete containing ground granulated blast furnace slag as cement replacement will be referred as slag concrete hereafter. Data was compiled from the studies that evaluated the salt scaling resistance of slag concrete by employing ASTM C672 (or MTO LS-412). The data is analyzed in two subsets: the first group contains the data based on visual rating and the second group consists of the data based on mass loss. The data set includes concretes that contain only portland cement and portland cement-slag combinations; are lab-produced; are 28 days cured; and are tested for 50 freeze-thaw cycles. Otherwise, the chemistry of the cementitious materials, mix proportions, w/cm, specimen preparation and surface treatment vary. Figure 1 shows the salt scaling assessment data that is based on visual rating and has been compiled from Luther et al. (1994), Duos and Eggers (1999), Soeda et al. (1999), Luther et al. (2000), Rivest et al. (2004) and Bouzoubaa et al. (2008). The plot contains 105 data points including a significant number of coincident points due to the nature of rating system. Figure 2 shows the data that is based on mass loss from the surface and contains 177 data points compiled from Bilodeau et al. (1987), Afrani et al. (1994), Talbot et al. (1996), Boyd and Hooton (1997), Soeda et al. (1999), Cramer and Walls (2001), Sakai et al. (2001), Bleszynski et al. (2002), Rivest et al. (2004), Chidiac et al. (2005), Cramer and Sippel (2005), Krishnan et al. (2006), Boyd and Hooton (2007) and Bouzoubaa et al. (2008). 69 2

5 5 4 Scaling Rating (0-5) Slag Content (%) FIGURE 1. Plot of ASTM C672 rating data versus slag content. A common approach to assessing the output from ASTM C672 results is to designate a scaling rating of 2 as slight to moderate scaling and this figure is taken as an acceptance limit for the purposes of this analysis. Other authorities reportedly use a visual rating of 3 as an acceptance limit. The ASTM visual rating is based on assessment of the specimen surface by an operator and is therefore subjective. The mass of material scaled off the surface may be considered a preferred measure and a maximum limit of 0.8kg/m 2 is imposed by the MTO. Figure 2 shows the published data reflecting mass loss values against slag content. 79 3

6 Mass Loss (kg/m 2 ) kg/m Slag Content (%) FIGURE 2 Plot of mass loss data versus slag content. There is no obvious trend line apparent in either figure. However, some interesting observations can be extracted from them as discussed below. It appears from these data that the probability of producing concrete that is durable under freeze thaw conditions is largely independent of slag content. This implies that there are various other factors that must be contributing to the poor performance of some of the systems. This is consistent with other published findings. In their study involving multi-variable statistical analysis of scaling resistance of slag concrete, Panesar and Chidiac (2007) concluded that the amount of slag do not singularly control the mass loss but there is an unclear interrelation between ordinary portland cement content, slag and total binder contents. Both the finishing procedure and timing reportedly have a significant effect on the scaling resistance for systems with and without supplementary cementitious materials (Hooton and Boyd (1997), Taylor et. al. (2004)). The chemistry of the portland cement also appears to influence the results of the scaling test made using similar mix proportions and with a narrow range of air contents (Taylor 2009b). The data were also grouped into four sets according to the slag content no slag (0%), 1-30%, 31-50%, and >50% and plotted based on the passing criteria a rating of 2 or a mass loss of 0.8 kg/m 2. The results are given in Figure 3. It appears that, particularly according to the mass loss criterion, the probability of a mixture performing poorly appears to be somewhat higher with slag contents above 30%, but there are still a significant proportion of systems performing well within the set. Above 50% it is less likely that a mixture will provide good resistance to salt scaling, although the amount of data at 4

7 these values is limited. However, based on the plots in Figures 1-3 it is also apparent that up to about 50% slag content it is readily possible to construct systems that pass the laboratory test Passing (%) >50 Slag (%) 103 ASTM Rating (2) Mass (0.8kg/m2) FIGURE 3 Scaling performance (Pass/Fail) based on ASTM rating and mass loss criteria. COMPARISON BETWEEN LABORATORY TEST RESULTS AND FIELD PERFORMANCE When considering the application of these data to specifications, it should also be noted that field performance of slag concrete generally does not align with the laboratory results. Luther et al. (2000) reported field test results from Weirton, West Virginia: after 6 years, even a mix containing 65% slag had showed no scaling. Boyd and Hooton (2007) published the results of in-ground slabs containing 25, 35, and 50% of slag. The slabs had been subjected to approximately 600 field freeze-thaw cycles over a 10- year period. Only certain sections of 50% slag concrete showed minor scaling which did not require any repair. In a recent study at ISU of 28 field sites, it was concluded that that construction-related issues played a bigger role in the observed scaling performance than did the amount of slag in the concrete mixture. For example, a pavement concrete from Delaware, with 50% slag experienced no scaling after 15 years of service (Schlorholtz and Hooton 2008). Besides the construction practices, the degree of hydration, or maturity, at the time concrete is exposed to deicer salts was reported to be important (Bouzoubaa et al. 2008). Chidiac and Panesar (2008) concluded that testing for scaling at the age of 28 days leads to lower scaling resistance for concrete containing high percentages of slag because of the limited curing regime of the ASTM procedure. However, field concrete is generally better hydrated before its exposure to deicer salts compared to laboratory concrete. Poor correlation between ASTM C672 results and field performance has also been reported for fly ash concrete (Thomas 1997). The rock salt used on concrete pavements for deicing purpose may highly vary in chemical composition and 5

8 perhaps not as pure as the salt used in laboratory testing. This is probably another factor contributing to variability in the salt scaling performance of laboratory and field concrete. Bouzoubaa et al. (2008) recommended that the BNQ NQ (BNQ: Bureau de Normalisation du Québec) procedure offers better reproducibility and better correlation with field performance than the ASTM C 672 test. A concrete mix used for sidewalk application that contained 35% slag showed no scaling in the field (rated as 0-1 on a visual scale). However the same mix was rated in the laboratory as 4 using the ASTM C672 procedure and 1 according to BNQ. In the same publication, the major differences between the ASTM and BNQ methods were described as: Finishing: The BNQ standard does not require brushing after bleeding has ended. The specimens are covered with a plastic sheet and a geotextile is put at the bottom of the mold to provide drainage. This reduces variability due to differences in finishing techniques and timing. Pre-saturation: The BNQ procedure uses a 13-day moist curing followed by 14-day drying and finally 7-day saturation of surface with a 3% NaCl solution thus reducing the effects of osmosis on the test. Salt Solution: 3% NaCl is the salt solution used and the samples are required to be covered to prevent evaporation and so concentration of the solution. Mass Loss: The scaling residue is collected after 7, 21, 35 and 56 freeze-thaw cycles. The limit is 0.5kg/m 2 after 56 cycles. CONCLUSION Based on observations drawn from a number of references describing field and laboratory work investigating scaling resistance of concrete the following conclusions may be drawn: Up to about 50% slag content it is possible to construct systems that pass the laboratory test. Above 50% it is less likely that a mixture will provide good resistance to salt scaling. The current ASTM test method for salt scaling, ASTM C 672, does not reliably predict performance of concrete mixtures in the field, particularly those containing slag. The test is more severe than the field exposure. An alternative test method that exhibits a better correlation with field performance is needed. In the field, construction-related issues probably play a bigger role in observed scaling performance of concrete mixtures than does the amount of slag. Other factors such as air content, finishing activities and cement chemistry contribute to the variability in mixture performance in the laboratory and the field. Based on these observations it appears reasonable to permit slag dosages up to 50% by mass of cement in concrete exposed to salt in a freezing environment. This is consistent with the requirements for structural concrete given in ACI 318. All cases, care must be taken to ensure proper mix proportions and workmanship to increase the likelihood that the mixture will perform satisfactorily. 6

9 REFERENCES Afrani, I., and C. Rogers. The Effects of Different Cementing Materials and Curing on Concrete Scaling. Cement, Concrete, and Aggregates, Vol. 16, No. 2, 1994, pp Bliodeau, A., G. G. Carette, V. M. Malhotra. Resistance of Concrete Incorporating Granulated Blast Furnace Slag to the Action of De-icing Salts. In Proceedings of the International Workshop on Granulated Blast-Furnace Slag in Concrete, Toronto, Ontario, Canada, 1987, pp Bleszynski, R. R., D. Hooton, M. D. A. Thomas, and C. A. Rogers. Durability of Ternary Blend Concrete with Silica Fume, and Blast-Furnace Slag: Laboratory and Outdoor Exposure Site Studies. ACI Materials Journal, Vol. 99, No. 5, 2002, pp Bouzoubaa, N., A. Bilodeau, B. Fournier, R. D. Hooton, R. Gagne, and M. Jolin. Deicing Salt Scaling Resistance of Concrete Incorporating Supplementary Cementing Materials: Laboratory and Field Test Data. Canadian Journal of Civil Engineering, Vol. 35, No. 11, 2008, pp Boyd, A. J., and R. D. Hooton. Effect of Finishing, Forming and Curing on De-icer Salt Scaling Resistance of Concretes. In Proceedings of the International RILEM Workshop on Resistance of Concrete to Freezing and Thawing With or Without De-icing Chemicals, Essen, Germany, 1997, pp Chidiac, S. E., D. K. Panesar, P. Smeltzer, and E. Kling. Current Issues with Concrete Containing Blast Furnace Slag in the Precast Instry. In Proceedings of the First Canadian Conference for Effective Design of Structures, Hamilton, Ontario, Canada, 2005, pp Chidiac, S. E., and D. K. Panesar. Evolution of Mechanical Properties of Concrete Containing Ground Granulated Blast Furnace Slag and Effects on the Scaling Resistance Test at 28 Days, Cement and Concrete Composites, Vol. 30, No. 2, 2008, pp Cramer, S. M., and R. A. Walls, Jr. Strategies for Enhancing the Freeze-Thaw Durability of Portland Cement Concrete Pavements. Publication WI-SPR Wisconsin Department of Transportation, Duos, C., and J. Eggers. Evaluation of Grade 120 Granulated Ground Blast Furnace Slag. Publication LTRC Project No. 96-3C. Louisiana Transportation Research Center, Federal Highway Administration, Ground Granulated Blast-Furnace Slag. Accessed July 15, Hooton, R. D., and A. Boyd. Long-Term Performance of Concretes Containing Supplementary Cementing Materials. Journal of Materials in Civil Engineering, Vol. 19, No. 10, 2007, pp Krishnan, A., J. K. Mehta, J. Olek, and W. J. Weiss. Technical Issues Related to the Use of Fly Ash and Slag During Late-Fall (Low Temperature) Construction Season. Publication FHWA/IN/JTRP-2005/5. Indiana Department of Transportation and Federal Highway Administration,

10 Luther, M. D., W. J. Mikols, A. J. DeMaio, and J. E. Whitlinger. Scaling Resistance of Granulated Blast Furnace (GGBF) Slag Concretes. In Proceedings of the Third CANMET/ACI International Conference on Durability of Concrete, Nice, France, 1994, pp Luther, M. D., D. J. Imse, and L. L. LaFallotte. Resistance to Deicer Scaling Concrete Containing Ground Granulated Blast Furnace. In Proceedings of the Fifth CANMET/ACI International Conference on Durability of Concrete, Barcelona, Spain, 2000, pp Marchand, J., E. J. Sellevold, and M. Pigeon. The Deicer Salt Scaling Deterioration of Concrete An Overview. In Proceedings of the Third CANMET/ACI International Conference on Durability of Concrete, Nice, France, 1994, pp Panesar, D.K., and S.E. Chidiac. Multi-variable Statistical Analysis for Scaling Resistance of Concrete Containing GGBFS, Cement and Concrete Composites, Vol. 29, No. 1, 2007, pp Rivest, M., N. Bouzoubaa, and V. M. Malhotra. Strength Development and Temperature Rise in High- Volume Fly Ash Concretes in Large Experimental Monoliths. In Proceedings of the Proceedings of the Eight CANMET/ACI International Conference on Fly Ash, Silica Fume, Slag and Natural Pozzolans in Concrete, Las Vegas, Nevada, USA, 2004, pp Sakai, K., M. Kumagai, K. Abe, and H. Endoh. Effect of Pore Structure on Scaling Deterioration of Concrete. In Proceedings of the Third International Conference on Concrete under Severe Conditions Environment and Loading, Vancouver, British Columbia, Canada, 2001, pp Schlorholtz, S., and R. D. Hooton. Deicer Scaling Resistance of Concrete Pavements, Bridge Decks, and Other Structures Containing Slag Cement, Phase 1: Site Selection and Analysis of Field Cores. Publication CTRE Project National Concrete Pavement Technology Center, Sippel, C. and S. Cramer. Effects of Ground Granulated Blast Furnace Slag in Portland Cement Concrete. Publication WHRP Wisconsin Department of Transportation, Soeda, M., T. Yamato, and Y. Emoto. Frost Durability of High-Performance Concrete Incorporating Slag or Silica Fume. In Proceedings of the Second International Conference on High Performance Concrete and Performance and Quality of Concrete Structures, Rio Grande do Sul, Brazil, 1999, pp Talbot, C., M. Pigeon, J. Marchand. Influence of Supplementary Cementing Materials on the De-icer Salt Scaling Resistance of Concrete. In Proceedings of the Seventh International Conference on Durability of Building Materials and Components, Stockholm, Sweden, 1996, pp Taylor, P. C., W. Morrison, and V. A. Jennings. The Effect of Finishing Practices on Performance of Concrete Containing Slag and Fly Ash as Measured by ASTM C 672 Resistance to Deicer Scaling Tests. Cement, Concrete, and Aggregates, Vol. 26, No. 2, 2004, pp

11 Taylor, P. C., W. A. Pyc, S. Y. Lee, and J. Z. Zemajtis. Effect of Carbonation on Deicer Resistance of Concrete Containing Fly Ash or Slag. Publication PCA R&D Serial No. SN2827. Portland Cement Association, 2009a. Taylor, P.C. Effect of Clinker Chemistry on Salt Scaling Resistance of Concrete. In Transportation Research Record: Journal of the Transportation Research Board, Concrete Materials, Transportation Research Board of the National Academies, Washington, D.C., 2009b, accepted for publication Thomas, M. D. A. Laboratory and Field Studies of Salt Scaling in Fly Ash Concrete. In Proceedings of the International RILEM Workshop on Resistance of Concrete to Freezing and Thawing With or Without Deicing Chemicals, Essen, Germany, 1997, pp Valenza II, J. J., and G. W. Scherer. A Review of Salt Scaling: I. Phenomenology. Cement and Concrete Research, Vol. 37, No. 7, 2007, pp