Effect of curing conditions on freeze-thaw de-icing salt resistance of blast furnace slag cement mortars

Transcription

1 Effect of curing conditions on freeze-thaw de-icing salt resistance of blast furnace slag cement mortars O. Çopuroğlu, A. L. A. Fraaij & J. M. J. M. Bijen Section of Materials Science and Sustainable Construction, Faculty of Civil Engineering and Geosciences, Delft University of Technology, the Netherlands Abstract The freeze-thaw de-icing salt scaling resistance tests were performed on blast furnace slag cement (BFSC) mortars, which were cured in 3% NaCl solution, tap water (with/without plastic bags), demineralised water, CaSO 4 saturated solution, Ca(OH) 2 saturated solution or NaOH saturated solutions. Additionally, the effect of carbonation on the scaling resistance was studied. The effect of carbonation on the microstructure of the BFSC mortar skin was investigated by optical microscopy. The results showed that de-icing salt scaling resistance of a BFSC mortar is strongly dependent on the occurrence of carbonation during curing regime and to a less extent on the chemical composition of curing water. Keywords: freeze-thaw scaling, curing, carbonation, Ca ion leaching, durability. 1 Introduction The Dutch cement and concrete market distinguishes itself with respect to the relatively high use of cements with fillers and ground granulated blast furnace slag. BFSC has a market share of more than 50%. The energy and primary raw material consumption for the production of this cement is particularly low in comparison to ordinary portland cement. The BFSC shows a favorable resistance to ingress of the aggressive compounds (notably chloride) and to chemical attack. However a disadvantage of BFSC is its generally low freeze-thaw salt resistance. With respect to ordinary Portland cement (OPC), it often appears that BFSC has a faster surface deterioration, which results in the early disappearance of the cement skin.

2 234 High Performance Structures and Materials II In the decision making process for infrastructural work, BFSC is often not chosen due to the relatively low resistance against freeze-thaw salt attack. This plays an important role in countries where freeze-thaw salt deterioration is an issue. In The Netherlands, BFSC is scarcely or not applied for structures, which are heavily influenced by freeze-thaw salts. The improvement of this behavior could stimulate the application of this cement. As a part of a PhD project, current study was performed in order to understand the causes of de-icing salt attack against BFSC mortars. This report presents the initial results of the experimental study concerning the difference in de-icing salt scaling behavior from curing conditions point of view. 2 Technical background BFSC concrete appears to be more sensitive to curing than OPC with the same workability. This is due to the lower hydration rate and increase in permeability when carbonated, where for OPC it alternates. Consequently replacement of cement by equal mass percentages of slag will generally make concrete more vulnerable to degradation by de-icing salts [1]. It was reported that when the concrete is designed with an identical strength development, the differences in freeze thaw de-icing salt resistance between OPC concrete and concrete with BFSC will be smaller. For instance, Wiebenga showed during the first cycles that BFSC concrete shows behavior similar to the concrete with OPC. According to him the differences between concrete with BFSC and OPC concrete would be expected to be smaller when curing is prolonged [2]. Apart from carbonation and hydration rate, the effect of the curing media is also believed to be important on salt scaling resistance of mortars and concrete. Stark and Ludwig studied the effect of the quality of curing water and unexpectedly found that the harder the curing water the higher the salt scaling of BFSC and OPC concretes [3]. 2.1 Carbonation While carbonation leads to a slight densification of the microstructure in OPC mortar, the microstructure of BFSC mortar becomes coarser. In fact, the total porosity decreases in BFSC mortar, while the portion of capillary pores increased. There are mainly two different reactions with related to carbonation. These are: Ca( OH ) 2 + CO2 CaCO3 + H 2O (vol. increase) (1) CSH + CO2 SiO2 nh 2O + CaCO3 + H 2O (vol. decrease) (2) As can be understood above, the resulting volume of the mortar depends on the amount of Ca(OH) 2 (portlandite) and C-S-H. Having less portlandite would lead BFSC concrete to be more porous.

3 High Performance Structures and Materials II Experimental Even though it seems that concrete testing is more practical for the site concreting, mortar testing is technically easier and economical, however yet nonstandardized. For the matter of comparison, mortars are also believed to be suitable according to many researchers [4, 5, 6]. In this report, two aspects of curing of cementitious materials are especially considered, which may have crucial effects on salt scaling resistance - carbonation and curing media. 3.1 Materials CEM III/B 42.5 ENCI BFSC, which had a slag content of 68% was used in the mortars used. Chemical and physical properties of the cement are given in table cm mortar bars were prepared according to EN196-1, keeping the water cement ratio in the order of 0.50 for all mortars. The mortar forms were vibrated for 40 seconds in total and filled in two layers. Air content of fresh mortar was measured as 1.5%. Table 1: Physical and chemical properties of CEM III/B Chemical Physical Strength of standard mortars CaO % days N/mm SO 3 % days N/mm SiO 2 % days N/mm Al 2 O 3 % 12.0 Blaine m 2 /kg Curing Eleven different curing media were selected. These were: 1. 3% (by mass) NaCl solution: This extreme situation was tested to simulate seawater at 20 C during 5 week curing. (Code: CL) 2. Sealed: Specimens were immersed in the tap water inside the plastic bags in order to provide minimum Ca leaching at 20 C for 5 weeks. Water was allowed to penetrate inside the bags (Code: Z) 3. Strongly dried, no carbonation (I): After initial 6 days of water curing, specimens were kept in the desiccator under the condition of 0%RH and 0% CO 2 at 20 C during 3 weeks prior to 3% NaCl saturation period. Silica gel was used for collecting the humidity from the environment. (Code: E) 4. Dried, no carbonation (II): After initial 6 days tap water curing, the samples were put in the desiccator for 3 weeks. Relative humidity was kept in the order of 50% by means of Mg(NO 3 ) 2.6H 2 O (Magnesium Nitrate). A blend of NaOH and Ca(OH) 2 (Sodalime pellets with absorption cap. 28% for CO 2 )

4 236 High Performance Structures and Materials II was used to collect CO 2 from air and to prevent carbonation. Curing program ended with 1 week of 3% NaCl solution saturation prior to freeze thaw salt cycles. All conditions were at 20 C (Code: AA) 5. Dried, carbonated (I): After initial 6 days water curing, specimens were kept in the laboratory under the condition of 50% R.H. and 20 C during 3 weeks prior to 3% NaCl saturation period. (Code: AB) 6. Dried, accelerated carbonated (II): After initial 6 days water curing, specimens were kept in the special container with 3% CO 2 under the condition of 50% R.H. and 20 C during 3 weeks prior to 3% NaCl saturation period. (Code: AC) 7. Demi water: Demineralised water was used at 20 C for 5 weeks continuous curing. (Code: DW) 8. Plain water: Tap water was used for this curing. In this method, cement mortars were kept in tap water at 20 C for 5 weeks and then they were exposed to freeze thaw salt cycles. (Code: LW) 9. CaSO 4 solution: Saturated CaSO 4 solution was used was also used at 20 C for 5 weeks. (Code: S) 10. Lime water: Saturated lime solution was used for curing at 20 C during 5 weeks. (Code: H) Molar NaOH solution: The idea of using this curing solution was to keep the curing media above ph 13 and increase the rate of slag activation. Solution was kept at 20 C (Code:N). For the carbonation series specimens (E, AA, AB, AC), initial water curing, drying period and water suction durations were kept as follows: (Tap) water curing 6 days Drying and carbonation period 21 days Water suction/ 3% NaCl solution saturation 7 days Other specimens were kept in the relevant media for 5 weeks continuously. In all conditions the temperature was at 20 C. 3.3 Freeze-thaw test Following the curing period, specimens were cut in two and subjected to freeze thaw cycles of 17±1 hrs freezing at -20±2ºC and 7±1 hrs thawing at 20±2 ºC. 14 cycles were completed for each series and scaled material was collected by filtration after 2, 7 and 14 cycles. Scaled materials were placed in a stove at 110ºC and weighed after 24 hrs drying. Before freeze thaw testing, lateral sides of the mortar bars were covered with two-layer isolation material, epoxy plus bituminous tape. Aim of this preparation was to fix the exposed volume and to prevent potential danger of additional scaled material falling from laterals of the specimen, which would lead to inaccurate measurement (see figure 1).

5 High Performance Structures and Materials II 237 Figure 1: Schematic description of a freeze-thaw sample. 3.4 Chemical analysis Curing water samples were collected each week during the entire curing period and fresh curing media was replaced with previous one. Chemical analysis of the curing water was carried out by atomic emission spectrometry (ICP AES) after 5 weeks just before freeze-thaw cycles. 3.5 Microstructural analysis DBT Dimond Roller and Grinder/86 thin sectioning machine was used for the preparation of the thin sections of the carbonated specimens. Low viscosity epoxy with special yellow color epo-dye was used for distinguishing the porosity under fluorescent light of optical microscope. Quantimet color image analyzer from Leica was used for image capturing and porosity measurements. Images of the samples were taken just before the freeze-thaw cycles began. 4 Results and discussion Strength data of the specimens are given in table 2. These results were measured after 5 weeks immediately before freeze-thaw cycles. Table 2: Average compressive strengths of the 5-week old mortars. Curing Environment Comp. Strength (MPa) Curing Environment Comp. Strength (MPa) CL DW Z LW E S AA H AB N AC Age of testing 35 days

6 238 High Performance Structures and Materials II Strength measurements showed that the higher the concentrations of CO 2 in air the lower the compressive strength of the BFSC mortars (see series E, AA, AB and AC with increasing CO 2 attack). The reason is most probably the increase in porosity and decrease in cross sectional area of the solid phase of the specimens due to carbonation and drying. Water cured specimens had higher and identical compressive strength values compared to dried series (E: extreme drying, no carbonation). Figure 2 summarizes the salt scaling performances of differently cured mortars cycles 7 cycles 14 cycles 8 7 Scaling (gr.) CL Z E AA AB AC DW LW S H N Curing Type Figure 2: Scaling performances of CEM III 42.5 /B cement mortars in different curing media. In order to see the effect of the concentration of CO 2 in air on de-icing salt scaling of BFSC mortars, a comparison should be made between dried - noncarbonated (codes E and AA) and dried carbonated (AB and AC) specimens. Since the effect of drying is eliminated, it can be concluded that under constant relative humidity carbonation has crucial effect on de-icing salt scaling of BFSC mortars. This is probably the consequence of newly formed pore structure of the skin part of the mortars. The effect of drying should be taken into account as well. The comparison between series E (0% CO 2, 0% RH) and series AA (0% CO 2, 50% RH) reveals that under non-carbonated condition, the salt scaling resistance of BFSC mortar decreases as the rate of drying increases. However, the effect of drying is much less significant than the effect of carbonation. In figure 3 the similarity of tendency of the scaling progresses of BFSC mortars and BFSC concrete are presented. The results of Stark and Ludwig for

7 High Performance Structures and Materials II 239 concrete were taken for the comparison with the BFSC mortar scaling results [7]. According to the results, in mortar testing approximately 3 grams of scaling corresponds to 1500 gr/cm 2 threshold of concrete testing. Figure 4 provides the micro images of the coarsening of the porosity of mortar skin due to carbonation. The differences in the paste and interfacial transition zone are noteworthy. Depth of carbonation of these series was also measured by phenolphthalein spraying method. The average values for series AA, AB and AC were 0, 1 and 10 mm respectively. Figure 3: (a) Effect of carbonation level on de-icing salt scaling performances of BFSC mortar. In all cases, relative humidity was 50% and slag content was 68%. (b) Effect of carbonation level on de-icing salt scaling performances of BFSC concrete. In all cases, relative humidity was 65% and slag content was >60% [7]. It was also investigated to see whether the better de-icing salt scaling performance of the mortars (cured in Ca(OH) 2 solution) would come from enhanced slag reaction due to availability of lime. Considering the results, it is possible to say that the Ca ion in the entire system is more essential for de-icing

8 240 High Performance Structures and Materials II salt resistance of the mortar than curing by NaOH solution, at least for the initial 5 weeks of curing period (see figure 1, comparison between N and H series). Nevertheless one should take into consideration that this better behavior of the mortars cured in Ca(OH) 2 solution could be the consequence of one of the following conditions; prevention of Ca ion leaching from the mortar, penetration of Ca ion into mortar so having more lime for slag hydration or precipitation of CH in the pores. For definite answer further research is needed. Figure 4: Effect of carbonation on the skin porosity of blast furnace slag cement mortar. (a) 3% CO 2, total porosity in skin approx. 20% in average b) lab air, total porosity in skin approx. 10% in average. (c) no carbonation, total porosity in skin approx 5% in average. Magnification is 25 times. Amount of Ca ion leaching after 5 weeks is given in table 3. Values are the net amount of leached ions from the specimens. Table 3: Ca ion content of curing water (mg/l). Series DW H CL S LW N Leached Ca ions Conclusion Considering the results, following conclusions can be drawn: 1. Carbonation is of paramount importance on freeze thaw salt scaling resistance of BFSC mortar phase. There is evidence that this is due to the restructuring of capillary system. 2. Freeze thaw de-icing salt scaling resistance of a BFSC mortar is dependant on the curing regime and chemical composition of curing water although to a much less extent compared to carbonation. 3. Curing in limewater appears to be favorable with respect to freeze thaw deicing salt scaling.

9 High Performance Structures and Materials II 241 References [1] Bijen, J.M.J.M, Blast Furnace Slag Cement for Durable Marine Structures, VNC/Betonprisma, The Netherlands, [2] Wiebenga, J., Vorstbestndigheid en Dooizoutbestandheid van Vliegcementbeton, TNO-IBBC-report B / , [3] Stark, J & Ludwig, H.-M, RILEM Proceedings 30 - Freeze Thaw Durability of Concrete, Ed. Marchand et. Al., E&FN Spon, pp , [4] Girodet, C., Chabannet, M., Bosc, J.L. & Pera, J., Influence of the type of cement on the freeze-thaw resistance of the mortar phase of concrete, Frost resistance of concrete, RILEM proceedings 34, Ed. M. Setzer and R. Auberg, E&FN Spon pp , [5] Browne, F.P. & Cady, P.D., Deicer Scaling Mechanisms in Concrete, ACI SP-47-6, American Concrete Institute, pp ,1976. [6] Peterson, O., Chemical Effects on Cement Mortar of Calcium Magnesium acetate as a De-icing Salt, Cement and. Concrete. Research., Vol.25, No.3, pp , [7] Stark, J. & Ludwig, H.-M., Freeze-deicing salt resistance of concretes containing blast-furnace slag-cement, Freeze Thaw Durability of Concrete, RILEM proceedings 10, Ed. J. Marchand et. Al, E&FN Spon, pp , 1997.