carte cement astm c 1280

PCA R&D Serial No. 2963

Effect of Portland Cement Fineness on ASTM C1260 Expansion

by F. Bektas, K. Wang, and H. Ceylan

©Portland Cement Association 2008 All rights reserved

1

KEYWORDS ASTM C 1260, alkali-silica reaction (ASR), ball mill, Blaine, expansion, portland cement, fineness ABSTRACT The present study is aimed at investigating the effect of portland cement fineness on the results of ASTM C1260 tests - Standard Test Method for Potential Alkali Reactivity of Aggregates (Mortar Bar Method). The effects of clinker alkali content, aggregate reactivity, and NaOH solution concentration on the mortar expansion test results were also studied. In this study, high and low alkali portland cement clinkers were selected and ground with gypsum in a laboratory ball mill to three fineness levels: 300, 400, and 500 m2/kg (Blaine). Moderately and highly reactive aggregates were tested with these cements according to ASTM C 1260. In addition to the standard 1 N NaOH solution, a 0.5 N NaOH soak solution was used. Furthermore, three commercially available portland cements (with different alkali content and fineness) were examined. The results show that mortar bar expansion was promoted with increased cement fineness regardless of clinker alkali, aggregate reactivity, or soak solution normality. Clinker alkali had little or no effect on moderately reactive aggregate, whereas it had considerable effect on highly reactive aggregate. Highly reactive aggregate tended to be more sensitive to cement fineness and alkalinity. The results from the commercially produced cements did not provide conclusive results due to the limited number of cements available. Cement fineness and clinker alkali content did not affect the decision on potential reactivity of the aggregates used in this study; however, they might change the decision for aggregate that falls close to the limit line. REFERENCE Bektas, F.; Wang, K., and Ceylan, H., Effect of Portland Cement Fineness on ASTM C 1260 Expansion, SN2963, Portland Cement Association, Skokie, Illinois, USA, 2008, 30 pages.

2

Effect of Portland Cement Fineness on ASTM C1260 Expansion

by Fatih Bektas*, Kejin Wang,** and Halil Ceylan***

1. INTRODUCTION 1.1. Background Since its initial discovery by Stanton in the early 1940s, alkali-silica reaction (ASR) has been one of the major durability concerns of the construction industry. ASR is the chemical interaction of concrete alkalies and reactive siliceous aggregate. The reaction results in a gel-like product which is capable of swelling in the presence of water. The swelling pressure is often sufficient to cause concrete cracking.

Detection of reactive aggregate prior to use in concrete is crucial in order to take action against the deleterious reaction. Numerous test methods have been proposed after six decades of research. Since ASR is a slow process, harmful effects in concrete may take years to exhibit, depending on favorable conditions. Therefore, almost all ASR tests involve acceleration of the reaction. To achieve this goal, at least one of the following conditions is aggravated:

• Alkali concentration • Temperature • Humidity

ASTM C 1260 – Standard Test Method for Potential Alkali Reactivity of Aggregates

(Mortar Bar Method), or more generically, the accelerated mortar bar test, has been intensively used all around the world under different codes which involve slight modifications. The test is based on the South African NBRI method proposed by Davis and Oberholster (1986). It has been very popular since it is relatively quick and easy to perform. The method requires the periodic length measurement of 25×25×285-mm mortar bars which are immersed in 1 N NaOH solution at 80°C. The length change, or expansion, after 14 days immersion (16 days since casting) is taken as the indication of potential reactivity. ASTM C1260 considers expansion of > 0.20% as reactive and < 0.10 as innocuous; expansion between 0.10% and 0.20% is inconclusive and requires additional testing. A similar CSA standard (CSA A23.2-25A) sets 0.15% expansion as the line between reactive and nonreactive material. On the other hand, the accelerated mortar bar test draws criticism as being very severe because some innocuous aggregates with proven field performance are found to be reactive. Therefore, it is recommended that an aggregate not be assessed as

* Graduate Student, Department of Civil, Construction and Environmental Engineering, Iowa State University, Ames, IA 50011. ** Associate Professor, Department of Civil, Construction and Environmental Engineering, Iowa State University, Ames, IA 50011. *** Assistant Professor, Department of Civil, Construction and Environmental Engineering, Iowa State University, Ames, IA 50011.

3

reactive solely based on ASTM C 1260 results (Berube and Fournier 1993, Folliard et al. 2006).

It is well-accepted that the hydration rate of cement increases with increasing cement fineness. This is almost true for any kind of chemical reaction: The reaction between a solid and a liquid or gas is promoted with increasing surface area of the solid. Similarly, when a finer cement is used, the increased rate of alkali release from the cement may aggravate the deleterious ASR of the concrete. Hobbs (1989) observed considerable variation in ASR expansion of mortar made with similar alkali contents and attributed this variation to the different rates of alkali release by the cement, among other factors. ASTM C 1260 requires the use of portland cement meeting ASTM C 150, which defines a lower limit of 280 m2/kg of fineness measured with the Blaine apparatus (ASTM C 204). However, with regard to the fineness, a wide variety of material can be used. According to a survey (Tennis 1998) investigating characteristics of portland cements produced in North America in the late 1990’s, the ranges of fineness are 310 to 497 m2/kg and 319 to 672 m2/kg for Type I and III, respectively. With this wide a range, the cements produced from the same clinker with fineness at the far ends of the scale may possess different properties, as also may be the case in the accelerated mortar bar test. Berra et al. (1998), in their study testing natural reactive sands with ASTM C 1260, concluded that the native alkali content and the specific surface area of the portland cement are significant factors affecting mortar bar expansions.

Although there have been studies concerning the effect of alkali content on ASTM C 1260 expansion, the effect of fineness has apparently never been given further attention. In this study, a detailed and controlled approach was undertaken to investigate the effect of portland cement fineness on ASTM C 1260 expansion in conjunction with other influential factors, namely alkali content of clinker, aggregate reactivity, and immersion solution concentration. 1.2. Objective and Approach The objectives of this project are to:

• Evaluate the effect of portland cement fineness on the expansion measured by the accelerated mortar bar method, ASTM C 1260

• Investigate the effect of other factors – alkali content of clinker, aggregate reactivity, and normality of immersion solution

To achieve the goals of the project, the following research activities were carried out:

• Production of portland cement with different fineness levels • Measurement of cement fineness by the Blaine method • ASTM C 1260 testing of reactive aggregates using the cements produced • Validation of findings using cements commercially available in the North American

market • Verification of aggregate reactivity employing scanning electron microscopy (SEM)

2. EXPERIMENTAL WORK 2.1. Materials Two types of clinker (low alkali and high alkali) were used in the study. The former was obtained from South Carolina, USA, and the latter from Ontario, Canada. Gypsum used to

4

produce portland cement was obtained from a source in Iowa, USA. Three commercially available portland cements (M1, M2, and M3) were also studied. The chemical compositions and Blaine fineness values of the clinkers and cements used are given in Table 1. Table 1. Chemical Composition of the Clinkers and Cements Used Clinker Commercially available cement

% High Alkali* Low Alkali* M1 M2 M3

CaO 63.5 64.6 64.1 61.1 61.5

Al2O3 5.7 5.0 4.9 4.4 4.3

Fe2O3 2.4 4.0 3.0 3.2 3.1

SiO2 20.4 20.9 20.6 19.3 19.5

MgO 2.4 1.2 1.8 4.2 4.1

SO3 3.3 2.9 2.7 4.1 3.8

Na2O 0.2 0.1 - 0.13 0.12

K2O 1.1 0.4 - 1.51 1.48

Na2Oeq 0.91 0.37 0.42 1.09 1.07 Loss on ignition 0.7 0.4 1.5 1.7 1.8

Fineness (m2/kg) Depending on grinding 365 401 583

C3S 62 64 67 68 69

C2S 12 11 8 4 4

C3A 11 7 8 6 6

C4AF 7 12 9 10 9 * Analyses of portland cement with 400 m2/kg Blaine fineness produced form the clinkers.

Two different alkali reactive fine aggregates were used in the study. Their origins are Nebraska, USA (denoted as NE) and New Mexico, USA (denoted as NM). The aggregates were selected according to their known ASR history obtained from personal communications (Tadros and Halsey 2005). NM is highly reactive, while NE can be considered moderately reactive. The aggregates were sieved and recombined in accordance with the grading requirements of ASTM C 1260. 2.2. Production of Portland Cements A custom-designed laboratory-type ball mill was purchased for the production of the portland cements. The ball mill has dimensions of 51 cm in diameter and 28 cm in length. It is made of carbon steel and equipped with a motor providing 30 RPM when it is loaded. The mill also has a casing for dust and noise control (see Fig. A.1 in Appendix). A dry grinding process was utilized.

Raw materials, clinker and gypsum, were crushed using a jaw-crusher and passed through a 2.36-mm (ASTM No. 8) sieve before final grinding. Both materials were dried in order to remove moisture which might lead to particles sticking to balls and the mill surface. Steel balls were used as grinding media. Ball charge was selected as 90 kg with equal weights of 37.5- and 25.0-mm spherical balls (Fig. A.2). A batch of 7.0-kg material (96.7% clinker and 4.3% gypsum by weight) was fed into the mill.

5

The aim of the laboratory ball mill grinding was to produce portland cement with controlled levels of fineness. To achieve this goal, the grinding curve, cement fineness versus grinding time, was first developed for each clinker. The cement fineness was determined in accordance with the Blaine method, ASTM C 204 – Standard Test Method for Fineness of Hydraulic Cement by Air Permeability Apparatus, which is widely accepted by cement manufacturers. Initially, sampling was done without taking the whole batch out of the mill; however, it was found that this created significant fluctuation in measured fineness. Therefore, the fineness was determined by taking all the material out so that the material sampling improved. Based on the grinding curve, portland cements with desired fineness levels were produced at a given grinding time and the actual fineness of the manufactured cement was also measured. 2.3. ASR Testing – ASTM C 1260 Mortar bars were prepared using the available reactive aggregates and the laboratory-produced cements. The bars were produced and tested in accordance with ASTM C 1260:

• The mortar was mixed in accordance with ASTM C 305 – Standard Practice for Mechanical Mixing of Hydraulic Cement Pastes and Mortars of Plastic Consistency.

• The bars were cast in steel molds which were covered with removable tapes and plastic wraps as release agent (Fig. A.3).

• The bars were demolded after 24 hours and cured for another 24 hours in water at 80°C.

• Initial reading was taken after demolding and zero reading was taken at the end of 24-hour hot water curing. Then the bars were transferred into 1 N NaOH solution. A 0.5 N solution was also used for the sample storage.

• The mortar bars were stored in air- and water-tight polypropylene containers (Fig. A.4) in a laboratory-type oven (Fig. A.5).

• The length change of the bars was recorded periodically up to 28 days instead of the 14 days recommended by the standard. The comparator used for the measurements is given in Fig. A6.

2.4. Scanning Electron Microscopy In order to corroborate the alkali reactivity of NE and NM aggregates, cross sections of 25×25 mm2 were cut from the mortar bars subjected to ASTM C 1260 conditioning for 28 days. The sections were cut using a low speed saw and the surface was polished using successive abrasive papers (down to 6 micron). Then, the sections were oven-dried at 40oC for 7 days before SEM operation gold sputtering was applied in order to achieve conductivity. Secondary electrons were used for imaging and energy dispersive spectroscopy (EDS) was utilized for qualitative chemical analysis. The images were taken at an accelerating voltage of 20kV. 2.5. Testing Parameters The experimental work includes two main phases – testing with laboratory-produced cements and testing with commercial cements for validation. Parameters of the first phase can be summarized as follows:

3 fineness levels: 300, 400, 500 m2/kg Blaine 2 clinkers: low alkali (~0.4% Na2Oeq) and higher alkali (~0.9% Na2Oeq)

6

2 reactive aggregates: low reactivity and high reactivity 2 different levels of NaOH solution normality: 1.0 N and 0.5 N

The second phase includes three commercially produced cements, allowing a

comparison between coarse and fine cement, and low and high alkali cement. 3. FINDINGS 3.1. Grinding Curve The grinding curve for the higher alkali clinker is given in Fig. 1, including all the trial batches and final products. Three fineness levels were targeted in the study – 300, 400, and 500 m2/kg. The cements are termed as H3, H4, and H5; H indicates higher alkali clinker and the preceding number represents the level of fineness, and similarly L3, L4, and L5 where L represents low alkali clinker. Desired fineness levels were achieved in a range of ±20 m2/kg. The Blaine values, together with the 45-µm (No. 325) sieve residues of the cements produced, are given in Table 2. The low alkali clinker was harder than the higher alkali clinker so that a longer grinding time was required to achieve the same fineness.

R2 = 0.98

R2 = 0.98

0

100

200

300

400

500

600

0 1 2 3 4 5

Grinding time (hours)

Bla

ine

finen

ess

(m2 /k

g)

High-alkali Low-alkali

Figure 1. Grinding curves of the high and low alkali clinkers. Table 2. Fineness of the Produced Portland Cements Grinding Time (minutes) Blaine Fineness (m2/kg) 45-µm Sieve Residue (%)

H3 47 306 23.9 H4 82 392 16.1 H5 155 505 12.0 L3 60 294 26.5 L4 120 403 21.2 L5 * 518 6.5

* Low alkali clinker could not be ground beyond 450 m2/kg Blaine. After 4 hours of grinding the material was sieved over a 75-μm (No. 200) sieve in order to remove bigger particles and achieve desired fineness.

7

3.2. Visual Observations and SEM Investigation At the end of testing period (extended to 28 days), mortar bars made with both NE and NM aggregate showed surface cracking regardless of the cement and solution concentration. The degree of cracking, in terms of surface area and crack width, was higher for the mortars made with the highly reactive NM aggregate than that of the mortars made with moderately reactive NE aggregate. Some of the mortar bars made with NM aggregate even exhibited warping (Fig. 2), which tends to be associated with high expansion.

Figure 2. Warping of the mortar bars incorporating highly reactive NM.

Microscopy of the sections of the bars made with M1 cement, proved the reactivity of the aggregates. Observations clearly demonstrated the internal deterioration caused by the deleterious ASR (Fig. 3). Crystallized ASR products were found in the voids. Both gels demonstrate rosette-like morphology, but with different arrangement of plate crystals (Figs. 4a and 5a). Rosette morphology is a typical crystalline variety of ASR gel observed previously (Regourd and Hornain, 1987). Figs. 4b and 5b show the composition of the observed gels. The EDS spectra clearly demonstrate the alkali-silica composition.

Figure 3. SEM micrograph of mortar bars cast with (a) moderately reactive aggregate and (b) highly reactive aggregate. Reacted particles and cracking can be identified at low magnification (R: Reacted particle).

(a) (b)

(a)

25x

R

R R

R

(b)

25x

R

R

R R

R

R

R

8

Figure 4. ASR gel product identified in mortar cast with moderately reactive aggregate and associated EDS spectrum.

Figure 5. ASR gel product identified in mortar cast with highly reactive aggregate and associated EDS spectrum. 3.3. ASTM C 1260 Expansion The ASR expansions of the mortar bars cast with the laboratory-produced cements are given in Figs. 6 to 9. Each data point represents the average of four bars. The data are given in Tables B.1 to B.8. The figures clearly indicate that there is an increase in the expansion as the fineness increases; the trend is valid regardless of clinker alkali, aggregate reactivity, and solution normality. However, the effect is more significant for the highly reactive aggregate.

Si

Ca

K Na

(a)

4000x

Si

Na Ca

K

4500x

(a) (b)

(b)

9

0.00

0.05

0.10

0.15

0.20

0.25

0 5 10 15 20 25 30

Time (Days)

Exp

ansi

on (%

leng

th c

hang

e)NE-H5-1N

NE-H4-1N

NE-H3-1N

NE-H5-0.5N

NE-H4-0.5N

NE-H3-0.5N

Figure 6. ASR expansion of less reactive aggregate with higher alkali cement.

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

0 5 10 15 20 25 30

Time (Days)

Exp

ansi

on (%

leng

th c

hang

e)

NM-H5-1N

NM-H4-1N

NM-H3-1N

NM-H5-0.5N

NM-H4-0.5N

NM-H3-0.5N

Figure 7. ASR expansion of highly reactive aggregate with higher alkali cement.

10

0.00

0.05

0.10

0.15

0.20

0.25

0 5 10 15 20 25 30

Time (Days)

Exp

ansi

on (%

leng

th c

hang

e)NE-L5-1N

NE-L4-1N

NE-L3-1N

NE-L5-0.5N

NE-L4-0.5N

NE-L3-0.5N

Figure 8. ASR expansion of less reactive aggregate with low alkali cement.

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

0 5 10 15 20 25 30

Time (Days)

Exp

ansi

on (%

leng

th c

hang

e)

NM-L5-1N

NM-L4-1N

NM-L3-1N

NM-L5-0.5N

NM-L4-0.5N

NM-L3-0.5N

Figure 9. ASR expansion of highly reactive aggregate with low alkali cement.

Fig. 6 gives the expansion curves of the moderately reactive NE aggregate tested in different solution concentrations with different fineness levels of the higher alkali cement, and Fig. 7 provides the same data for highly reactive NM aggregate. The effect of solution normality is worth mentioning since the ultimate expansions are comparable for the moderately reactive aggregate and even higher for the highly reactive NM aggregate. However, the incubation period of the reaction is extended in the solution with low normality. The trend is almost the same for the mortar bars cast with the low alkali cements. Figs. 10 and 11 demonstrate the measurements of bars cast with the cements obtained from the manufacturers. The expansion values are also tabulated in Tables B.9 to B.12. In this series, NM aggregate was tested as received in terms of gradation. Due to insufficient amount of material, the gradation of the NM aggregate was modified. The gradation is given

11

in Appendix C. M1 represents low alkali, low fineness cement; M2 represents higher alkali, low fineness, and similarly, M3 corresponds to the higher alkali, high fineness combination. Figs. 10 and 11 show the results for low and highly reactive aggregates, respectively. In each figure, (a) compares the fineness effect with similar alkalinity, and similarly, (b) compares the alkalinity effect with similar fineness cements. Fig. 10a shows that the finer cement, M3, leads to higher expansion regardless of the solution normality. However, the same trend could not be observed for the highly reactive aggregate as shown in Fig. 11a. In 1 N solution, the expansions are almost the same and in 0.5 N solution, M2, which is coarser compared to M3, produced higher expansion. For the alkalinity effect, M1 and M2, which have similar fineness, are compared; M1 has lower alkali content compared to M2. Fig. 10b shows the unexpected result that the low alkali cement increased the expansion regardless of the solution normality. On the other hand, the results of highly reactive aggregate shown in Fig. 11b outline different behavior. In 1 N solution, higher alkalinity leads to higher expansion, whereas this trend cannot be confirmed in 0.5 N solution. At 14 days, low alkali cement shows higher expansion and at 28 days the expansions are almost the same.

The study with the commercial cements did not produce consistent trends. Perhaps it would be more beneficial to carry out the tests with a larger number of cements representing a wider range of parameters under investigation.

12

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0 5 10 15 20 25 30

Time (Days)

Exp

ansi

on (%

leng

th c

hang

e)NE-M2-1N

NE-M3-1N

NE-M2-0.5N

NE-M3-0.5N

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0 5 10 15 20 25 30

Time (Days)

Exp

ansi

on (%

leng

th c

hang

e)

NE-M1-1N

NE-M2-1N

NE-M1-0.5N

NE-M2-0.5N

Figure 10. ASR expansion of moderately reactive aggregate with the commercial cements (a) similar alkalinity, different fineness (b) similar fineness, different alkalinity (see Table 1 for values).

(a)

(b)

13

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

0 5 10 15 20 25 30

Time (Days)

Exp

ansi

on (%

leng

th c

hang

e)

NM-M2-1N

NM-M3-1N

NM-M2-0.5N

NM-M3-0.5N

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

0 5 10 15 20 25 30

Time (Days)

Exp

ansi

on (%

leng

th c

hang

e)

NM-M1-1N

NM-M2-1N

NM-M1-0.5N

NM-M2-0.5N

Figure 11. ASR expansion of highly reactive aggregate with the commercial cements (a) similar alkalinity, different fineness (b) similar fineness, different alkalinity (see Table 1 for values). 4. DISCUSSION Fig. 12 summarizes the effect of parameters on the 14-day expansion. The effect of fineness is clearly shown in the figure, particularly for the highly reactive aggregate. The expansion is almost linearly correlated with the cement fineness. Fournier and Berube (1991) also observed higher expansion in the accelerated mortar bar method with the highest Blaine values. In their study with different clinkers, Berra et al. (1998) reported the exacerbating effect of fineness on the ASTM C 1260 expansion. Their results also indicate that the fineness effect is limited when using moderately reactive aggregate, which agrees with the results of this study.

(a)

(b)

14

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

200 300 400 500 600

Blaine fineness (kg/m2)

Exp

ansi

on a

t 14

days

(%)

H-NM

L-NM

H-NE

L-NE

Figure 12. Fourteen-day expansions of the mortar bars cast with high and low alkali clinker cements (NM: highly reactive sand; NE: moderately reactive sand).

One speculation is that finer cement may release alkalis faster, but little real investigation has been reported on this. In the present study, a simple pH test was performed for cements with different alkali levels and different fineness. The results are presented in Appendix D. Probably due to the limitation of the test method and testing time, the effect of cement fineness on the rate of alkali release is not clearly demonstrated by the present pH test results. Other test methods and longer testing time may be considered in future studies.

Fig. 12 shows that the effect of cement alkali content on the expansion is negligible for the moderately reactive aggregate. However, it is significant for the highly reactive aggregate; for the same fineness, higher alkali cement produced higher expansion. Although it is commonly accepted that the initial alkali level of the mixture has no significant effect on expansion because specimens are immersed in 1 N NaOH solution, contradictory data is also available in the literature. Berra et al. (1998) claimed that the alkali content of cement has an effect on ASTM C 1260 expansion. Owsiak (2003) reported parallel findings: Mixtures with 4% reactive opal produced 14-day expansions of 0.3%, 0.6%, and 0.7% with an initial mixture Na2Oeq of 0.35%, 0.70%, and 1.10% by cement weight. Berra et al. (1998) explained that there is a threshold level for alkali concentration in the pore solution within the mortar, above which the expansivity of the aggregate is promoted and sustained. This threshold exists even for the mortar bars immersed in NaOH solution and is affected by the fineness and alkali content of the portland cement, as well as by the permeability of the mortars.

The results of this study showed that the cements did not affect the classification of the aggregates tested as deleterious or innocuous. However, different fineness and different alkali content cements are capable of producing expansions in a band for the same aggregate (Figs. 13 and 14). This may lead to false decisions for certain aggregates, especially when the difference between reactive and innocuous is defined with a line as in CSA standards. In order to eliminate this situation, a buffer zone as in ASTM is useful.

15

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0 5 10 15 20 25 30

Time (Days)

Exp

ansi

on (%

leng

th c

hang

e)NE-H5-1NNE-H4-1NNE-H3-1NNE-L5-1NNE-L4-1NNE-L3-1N

Figure 13. Expansion curves of the moderately reactive aggregate in 1N NaOH.

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0 5 10 15 20 25 30

Time (Days)

Exp

ansi

on (%

leng

th c

hang

e)

NM-H5-1NNM-H4-1NNM-H3-1NNM-L5-1NNM-L4-1NNM-L3-1N

Figure 14. Expansion curves of the highly reactive aggregate in 1N NaOH. The cement fineness also has an effect on expansion for the mortar bars immersed in low solution concentration, 0.5 N. However, it is worth noting that the comparable expansion values observed in the low normality solution are not in agreement with the previous findings. On the other hand, ASR expansion is the result of a complex mechanism which is dependent on sensitive chemical equilibriums that have not been clearly explained. One case may not match another. For instance, Davies and Oberholster (1986) proposed 1 N solution concentration as the optimum since it produced higher expansion than 1.5 N. However, Fournier and Berube (1991) reported higher expansions for 1.5 N. Furthermore, ASTM C 1260 measures expansion depending solely on linear length change; however, in reality expansion is multidirectional. As shown in Fig. 2, some aggregates experienced warping, which may affect the measurements expressed in one direction. ASTM C 227, which also

16

involves the length measurements of mortar bars for ASR, recommends the documentation of warping. The same data might be a helpful inclusion for ASTM C 1260. 5. CONCLUSIONS Portand cements with different fineness levels, namely 300, 400, and 500 m2/kg, were produced in the laboratory using high and low alkali clinker. These cements were used to test moderately and highly alkali reactive aggregates in 0.5 N and 1 N NaOH solution following the ASTM C 1260 test procedure. A second series of mortar bars using commercially available portland cements was also investigated. The following conclusions are drawn based on the test results:

• As the cement fineness increased, the ASTM C 1260 expansion also increased regardless of aggregate reactivity, clinker alkali content, and NaOH solution concentration. The effect is more significant for the highly reactive aggregate.

• Cement alkali content did not significantly affect the expansion of the mortar made with the moderately reactive aggregate but tended to have significant effect on the expansion of the mortar made with the highly reactive aggregate.

• Cement fineness and alkali content did not affect the decision on potential reactivity of the aggregates used in this study; however, they might change the decision for aggregate that falls near a limit line.

ACKNOWLEDGEMENTS The research in this paper (PCA R&D SN2963) was conducted by Iowa State University with the sponsorship of the Portland Cement Association (PCA Project Index No. F05-05). The contents of this report reflect the views of the authors, who are responsible for the facts and accuracy of the data presented. The contents do not necessarily reflect the views of the Portland Cement Association.

The authors would like to express their great appreciation to the program managers Steven Kosmatka (PCA) and Beatrix Kerkhoff (PCA) for their support of this research, and Steve Otto (Holcim, US), Paul Tennis (Consultant), Tim Conway (Holcim, US), Nick Popoff (St. Marys, US), Bock Randall (St. Marys, US), Kimberley Kurtis (Georgia Institute of Technology), Lieska Halsey (Nebraska Department of Roads), Maher Tadros (University of Nebraska), and John Galvin (LaFarge, US) for their help in providing cement and aggregate materials, and Connie Field (PCA) for the literature support. REFERENCES Berube, M.A., and Fournier, B., “Canadian Experience with Testing for Alkali-Aggregate Reactivity in Concrete,” Cement and Concrete Composites, 15, 1993, pages 27 to 47. Berra, M.; Mangialardi, T., and Paolini A.E., “Testing Natural Sands for the Alkali Reactivity with the ASTM C 1260 Mortar Bar Expansion Method,” Journal of the Ceramic Society of Japan, 106, 1998, pages 237 to 241. Folliard, K.J.; Thomas, M.D.A., and Kurtis, K.E., Guidelines for the Use of Lithium to Mitigate or Prevent Alkali-Silica Reaction (ASR), Publication No. FHWA-RD-03-047, July 2006, 86 pages.

17

Fournier, B., and Berube, M.A., “Application of the NBRI Accelerated Mortar Bar Test to Siliceous Carbonate Aggregates Produced in the St. Lawrence Lowlands (Quebec, Canada). Part 1: Influence of Various Parameters on the Test Results,” Cement and Concrete Research, 21, 1991, pages 853 to 862. Hobbs, D.W., Alkali-Silica Reaction in Concrete, Thomas Telford Ltd., London, UK, 183 pages (p. 36). Oberholster, R.E., and Davies, G., “An Accelerated Method for Testing the Potential Alkali-Reactivity of Siliceous Aggregates,” Cement and Concrete Research, 16, 1986, pages 181 to 189. Owsiak, Z., ”Microstructure of Alkali-Silica Reaction Products in Conventional and Standard Accelerated Testing,” Ceramics – Silikaty, 47, 2003, pages 108 to 115. Regourd, M., and Hornain, H., “Microstructure of Reaction Products,” Proceedings of the 7th International Conference on Alkali-Aggregate Reaction in Concrete (Ottawa, 1986), Noyes Publication, 1987, pages 375 to 380. Tadros, M. (University of Nebraska) and Halsey, L. (Nebraska Department of Roads), personal communication, 2005. Tennis, P., “Portland Cement Characteristics – 1998”, Portland Cement Association Concrete Technology Today (PL992), 20, pages 1 to 3.

1-A

APPENDIX A

Equipment for Testing

2-A

Figure A.1. Custom designed laboratory type ball mill.

Figure A.2. Steel balls for grinding of clinker.

3-A

Figure A.3. Molds with removable tape and plastic wrap.

Figure A.4. Air-tight containers for storage of mortar bars in NaOH solution.

4-A

Figure A.5. Laboratory-type oven used to maintain 80oC.

Figure A.6. Comparator to determine the length change of mortar bars.

1-B

APPENDIX B

Mortar Bar Expansions

2-B

Table B.1. Moderately Reactive NE Aggregate with Higher Alkali Cement in 1N NaOH H3 H4 H5

Expansion(%) Std. Deviation Expansion(%) Std.

Deviation Expansion(%) Std. Deviation

0 0.00 0.00 0.00 3 0.03 0.00 0.03 0.00 0.04 0.00 7 0.08 0.00 0.09 0.00 0.08 0.00 10 0.10 0.00 0.11 0.00 0.12 0.01 14 0.12 0.01 0.12 0.01 0.14 0.00 21 0.14 0.01 0.15 0.01 0.19 0.01 28 0.16 0.01 0.17 0.01 0.20 0.00

Table B.2. Highly Reactive NM Aggregate with Higher Alkali Cement in 1N NaOH

H3 H4 H5



0 0.00 0.00 0.00 3 0.17 0.01 0.17 0.00 0.17 0.01 7 0.41 0.02 0.44 0.01 0.45 0.01 10 0.51 0.03 0.55 0.00 0.59 0.01 14 0.62 0.04 0.66 0.01 0.73 0.01 21 0.72 0.04 0.78 0.01 0.84 0.00 28 0.78 0.04 0.85 0.00 0.89 0.02

Table B.3. Moderately Reactive NE Aggregate with Higher Alkali Cement in 0.5N NaOH

H3 H4 H5



0 0.00 0.00 0.00 3 0.01 0.00 0.01 0.00 0.01 0.00 7 0.03 0.00 0.04 0.01 0.04 0.00 10 0.06 0.01 0.06 0.01 0.06 0.00 14 0.09 0.01 0.10 0.00 0.10 0.01 21 0.12 0.02 0.14 0.00 0.14 0.00 28 0.14 0.02 0.16 0.01 0.17 0.00

Table B.4. Highly Reactive NM Aggregate with Higher Alkali Cement in 0.5N NaOH

H3 H4 H5



0 0.00 0.00 0.00 3 0.06 0.01 0.09 0.01 0.11 0.01 7 0.36 0.02 0.43 0.01 0.43 0.03 10 0.56 0.03 0.58 0.05 0.63 0.05 14 0.72 0.03 0.80 0.01 0.81 0.05 21 1.00 0.04 1.04 0.02 1.08 0.06 28 1.12 0.07 1.17 0.03 1.25 0.05

3-B

Table B.5. Moderately Reactive NE Aggregate with Low Alkali Cement in 1N NaOH H3 H4 H5



0 0.00 0.00 0.00 3 0.03 0.01 0.03 0.00 0.03 0.01 7 0.08 0.00 0.10 0.01 0.10 0.00 10 0.10 0.01 0.12 0.01 0.12 0.00 14 0.12 0.01 0.13 0.01 0.14 0.00 21 0.14 0.01 0.15 0.01 0.16 0.00 28 0.16 0.01 0.16 0.01 0.19 0.00

Table B.6. Highly Reactive NM Aggregate with Low Alkali Cement in 1N NaOH

H3 H4 H5



0 0.00 0.00 0.00 3 0.14 0.01 0.15 0.00 0.18 0.01 7 0.41 0.02 0.44 0.01 0.44 0.02 10 0.50 0.00 0.55 0.01 0.57 0.02 14 0.55 0.01 0.61 0.01 0.68 0.01 21 0.61 0.00 0.68 0.01 0.76 0.01 28 0.65 0.01 0.73 0.01 0.82 0.01

Table B7. Moderately Reactive NE Aggregate with Low Alkali Cement in 0.5N NaOH

H3 H4 H5



0 0.00 0.00 0.00 3 0.01 0.00 0.01 0.00 0.02 0.00 7 0.03 0.00 0.03 0.00 0.04 0.00 10 0.07 0.01 0.07 0.01 0.10 0.00 14 0.10 0.00 0.11 0.02 0.14 0.01 21 0.12 0.01 0.12 0.01 0.17 0.01 28 0.14 0.01 0.14 0.02 0.18 0.01

Table B8. Highly Reactive NM Aggregate with Low Alkali Cement in 0.5N NaOH

H3 H4 H5



0 0.00 0.00 0.00 3 0.02 0.00 0.02 0.00 0.03 0.00 7 0.31 0.01 0.30 0.03 0.35 0.00 10 0.54 0.02 0.56 0.03 0.60 0.04 14 0.79 0.01 0.81 0.03 0.89 0.01 21 0.99 0.01 1.05 0.01 1.17 0.02 28 1.08 0.01 1.14 0.01 1.28 0.02

4-B

Table B.9. Moderately Reactive NE Aggregate with Commercial Cements in 1N NaOH

M1 M2 M3



0 0.00 0.00 0.00 4 0.07 0.00 0.04 0.00 0.04 0.00 7 0.18 0.01 0.10 0.01 0.09 0.01 10 0.20 0.01 0.12 0.01 0.14 0.01 14 0.22 0.01 0.16 0.01 0.18 0.02 21 0.29 0.03 0.20 0.01 0.23 0.01 28 0.31 0.01 0.26 0.01 0.30 0.01

Table B.10. Highly Reactive NM Aggregate with Commercial Cements in 1N NaOH

M1 M2 M3



0 0.00 0.00 0.00 4 0.33 0.01 0.42 0.01 0.36 0.01 7 0.76 0.01 0.75 0.03 0.72 0.01 10 0.89 0.00 0.94 0.07 0.89 0.02 14 0.96 0.01 1.03 0.03 1.05 0.03 21 1.06 0.01 1.17 0.04 1.20 0.04 28 1.12 0.02 1.26 0.04 1.31 0.04

Table B.11. Moderately Reactive NE Aggregate with Commercial Cements in 0.5N NaOH

M1 M2 M3



0 0.00 0.00 0.00 4 0.02 0.00 0.02 0.00 0.02 0.00 7 0.10 0.01 0.05 0.01 0.05 0.01 10 0.15 0.01 0.09 0.01 0.08 0.01 14 0.18 0.01 0.14 0.01 0.13 0.01 21 0.22 0.00 0.20 0.01 0.22 0.01 28 0.25 0.01 0.25 0.02 0.26 0.01

Table B.12. Highly Reactive NM Aggregate with Commercial Cements in 0.5N NaOH

M1 M2 M3



0 0.00 0.00 0.00 4 0.11 0.01 0.21 0.01 0.14 0.01 7 0.72 0.03 0.67 0.02 0.56 0.00 10 1.04 0.04 0.90 0.02 0.77 0.00 14 1.26 0.05 1.16 0.03 1.04 0.02 21 1.48 0.04 1.46 0.05 1.37 0.03 28 1.57 0.05 1.56 0.05 1.52 0.06

1-C

- APPENDIX C -

Aggregate Gradation

2-C

Table C.1. NM Aggregate Gradation Used in Testing of Commercial Cements

Sieve Size Mass, % Passing Retained on ASTM C 1260 Modified NM

4.75mm (No. 4) 2.36mm (No. 8) 10 14 2.36mm (No. 8) 1.18mm (No. 16) 25 5

1.18mm (No. 16) 600µm (No. 30) 25 19 600µm (No. 30) 300µm (No. 50) 25 39 300µm (No. 50) 150µm (No. 100) 15 23

1-D

- APPENDIX D –

pH Value of Ground Cement

2-D

In this measurement, 7 mg of cement was mixed with 7 ml of water in a PVC vial. A steel ball was placed inside the vial and the vial was rigorously shaken, employing a mechanical shaker. After 5 and 30 minutes of shaking, the pH value of the paste was measured by a pH meter, and the results are presented in Table D-1. (The test did not go beyond 30 minutes since the vials began to develop cracks.) Table D-1: pH Values of Ground Cements

Sample pH

5 minutes 30 minutes

L3 12.06 12.31

L5 12.00 12.27

H3 12.25 12.46

H5 12.26 12.47

In Table D-1, L and H denote alkali levels of the samples as low and high, respectively, and the following numbers represent the Blaine values 300 and 500 m2/kg. As seen in the table, there is a very small difference between the low and high alkali cements, which is hardly adequate to conclude the effect of cement fineness on the rate of alkali release. More investigations are needed but beyond our project funding and time budget limits. These investigations may include:

• Longer times for pH measurement • Extraction of pore solution • Measurements of soluble alkali concentrations and pH of the pore solution

carte cement astm c 1280

Documents