A STUDY OF PERMEABILITY AND TORTUOSITY OF CONCRETE

Shamsad Ahmad*, King Fahd University of Petroleum & Minerals, Saudi Arabia Abul Kalam Azad, King Fahd University of Petroleum & Minerals, Saudi Arabia Kevin F. Loughlin, King Fahd University of Petroleum & Minerals, Saudi Arabia

30th Conference on OUR WORLD IN CONCRETE & STRUCTURES: 23 - 24 August 2005, Singapore

Article Online Id: 100030015 The online version of this article can be found at: http://cipremier.com/100030015

Thisarticleisbroughttoyouwiththesupportof SingaporeConcreteInstitute www.scinst.org.sg AllRightsreservedforCIPremierPTELTD YouarenotAllowedtoredistributeorresalethearticleinanyformatwithoutwrittenapprovalof CIPremierPTELTD VisitOurWebsiteformoreinformation www.cipremier.com

30th Conference on OUR WORLD IN CONCRETE & STRUCTURES: 23 24 August 2005, Singapore

A STUDY OF PERMEABILITY AND TORTUOSITY OF CONCRETE

Shamsad Ahmad*, King Fahd University of Petroleum & Minerals, Saudi Arabia Abul Kalam Azad, King Fahd University of Petroleum & Minerals, Saudi Arabia Kevin F. Loughlin, King Fahd University of Petroleum & Minerals, Saudi Arabia

Abstract Permeability of concrete is commonly used as a key index for assessing the durability of concrete, especially when concrete is to be exposed to the aggressive environment. For a porous material like concrete, permeability depends on porosity and pore characteristics such as pore size, orientation, connectivity, and size variation. The effects of porosity and pore characteristics on permeability can be captured through a single parameter called tortuosity. Thus, tortuosity can also be viewed as an alternative index for concrete durability. In the present study, laboratory tests were carried out to measure permeability and tortuosity for different concrete mixes. A set-up, developed on the principle of Darcy's equation, was used for measuring permeability. Gas diffusion tests, PHI-220 Helium Porosimetry tests and Mercury Intrusion Porosimetry (MIP) tests were conducted for determining the tortuosity. Test results were used to study the variations of permeability and tortuosity with the key mix parameters such as water/cement ratio and cement content. A plot of tortuosity versus permeability values has indicated the existence of a typical relationship between tortuosity and permeability. Results show that higher the tortuosity, lower is the permeability.
Keywords: Concrete, porous material, permeability, tortuosity Introduction Since cement paste envelopes the aggregate particles in hardened concrete, it is the permeability of the paste that has the greatest influence on the permeability of concrete [1]. Permeability of the hydrated cement paste is closely related to the capillary porosity, which depends mainly on the solid/space ratio of paste. At a given degree of hydration, porosity increases with decrease in the solid/space ratio of paste. Therefore, the solid/space ratio, which is inversely proportional to the water/cement ratio, determines the permeability. The permeability steeply increases with decrease in the solid/space ratio below a critical value of 0.7 [2]. For the same degree of hydration, introducing sand particles into the cement paste, results in higher permeability [3]. However, it is widely reported that there is no defined relationship between porosity and permeability. Large variations in permeability of concrete having nominally similar porosities are frequently observed [4, 5, 6, and 7]. Permeability of concrete is not so much a function of the porosity as it is a function of the size, distribution and the continuity of capillary pores. High quality concrete at an early age has a total void volume of about 20 to 25% and about 10 to 25% at a mature age [8, 9]. Mehta and Manmohan [5] observed that the permeability of hardened cement paste is negligible (irrespective of its porosity) 1.

unless a network of micropores wider than a specific diameter of about 100 nm exists in the system. The volume of pores in most natural aggregates is usually under 3% and rarely exceeds 10% [2], while the capillary pores in a typical cement paste range from 20 to 40%. Many researchers have attempted to develop the applicable relationship between permeability and pore pressure since Powers et al. [10] first reported the relations between permeability and porosity. Nayme and Illston [4] reported an influence of maximum continuous pore radius on permeability. Hughes [6] reported the influence of average pore radius on permeability by using the Poiseuille formula. Mehta and Monteiro [2] took pore size distribution into account and proposed an empirical formula to express the relationship between permeability and pore structure. A model relating permeability of cement paste with its pore structure, determined by MIP, has been reported by Cui and Cahyadi [11]. Kumar and Bhattacharjee [12] have drawn the following conclusions from their research work on assessment of permeation quality of concrete through MIP [Kumar and Bhattacharjee, 2004]: (i) the permeation quality of concrete can be assessed on the basis of the knowledge of porosity and pore system characteristics of concrete such as equivalent pore radius and mean distribution pore radius of the concrete obtained through MIP results; (ii) the relationship involving mean distribution pore radius of the pores yields a better correlation than that involving equivalent pore radius; (iii) MIP can be used as an additional method for assessing the permeation quality of concrete. As a fluid flows essentially through the interconnected pore path in a porous material such as concrete, the flow path is tortuous in nature. It is assumed that all effects arising from pore orientation, connectivity, size variation, etc. can be lumped together as a single tortuosity factor. This is obviously a gross assumption since it depends on the implicit assumption that the effects of pore geometry and structure are the same for all pore sizes. Since the tortuosity factor is a structural parameter, it should not depend on the sorbate or the experimental conditions [13]. Dullien [14] has also shown that if the pore structure is characterized in sufficient detail, a reasonably accurate prediction of the tortuosity factor can be made. However, this requires a detailed knowledge of the pore shape as well as pore size distribution. In practice, it is generally simpler to treat the tortuosity as an empirical constant [13]. Tortuosity is usually defined as the square of the ratio of the average flow path length Le to the length L, along the major flow axis of the porous medium. Since in general Le > L, = (Le/L)2 > 1. However, Bear [15] defined tortuosity as (L/Le)2< 1. This is only a matter of definition Bear [15]. Values of tortuosity greater than 3, which are not uncommon [16, 17], are generally ascribed to the effect of variation in pore diameter which have been shown to cause a reduction in flux [18, 19, 20]. Bhatia [21, 22] has suggested that higher tortuosity may be attributed to a correlation effect which occurs even in a random lattice of uniform pores. In a recent study, an attempt has been made to use gas diffusion technique in measuring diffusion coefficient for concrete through the measurement of tortuosity [23]. Tortuosity,, porosity, , and diffusion coefficient, D, of a porous material (like mortar or concrete), are related to the pure gas diffusion coefficient, D0, as follows: = (D0/D) [24, 25]. Therefore, tortuosity can be determined through the measurements of diffusivity and porosity. In view of the fact that both permeability and tortuosity depend upon the pore structure and its interconnectivity, it is apparent that a relationship between these two material properties should exist for a porous medium [15]. This postulation has led to numerous attempts to correlate porosity and tortuosity to permeability. In the present work a total of nine mixes having different water/cement ratio and cement content were considered. Permeability and tortuosity for all the mixes were determined to study the effect of key mix parameters on permeability and tortuosity and also to observe the relationship between permeability and tortuosity.

2.

Experimental Investigation 2.1 Concrete Mixes

A total of 9 concrete mixes corresponding to 3 water/cement ratios (0.4, 0.5, and 0.6) and 3 cement contents (300, 350, and 400 kg/m3) at a constant CA/FA ratio 0f 1.6, as detailed in Table 1, were studied. Table 1: Details of the concrete mixes Mix proportions (by wt.) Cement : FA : CA 1 : 2.67 : 4.27 1 : 2.21 : 3.53 1 : 1.87 : 2.98 1 : 2.63 : 4.21 1 : 2.17 : 3.47 1 : 1.83 : 2.92 1 : 2.59 : 4.14 1 : 2.13 : 3.41 1 : 1.79 : 2.86

Mix No. 1 2 3 4 5 6 7 8 9

CA/FA ratio (by wt.) 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6

w/c ratio (by wt.) 0.4 0.4 0.4 0.5 0.5 0.5 0.6 0.6 0.6

Cement content (kg/m3) 300 350 400 300 350 400 300 350 400

ASTM C 150 Type I Portland cement was used in this study. The coarse aggregates for this study, was crushed limestone. The average values of specific gravity and absorption of such coarse aggregates, determined in accordance with ASTM C 127, are in the neighborhood of 2.5 and 1.5, respectively. The particle size distribution of coarse aggregate (by weight) was as follows: 15%, 70%, and 15% retained on sieve sizes 3/8 in., 3/16 in., and 3/32 inch, respectively. Medium coarse sand was used as fine aggregate. The specific gravity and absorption of the fine aggregates are typically found to be 2.6 and 0.57, respectively. For mixing the constituents, potable water was used. Moist curing was carried out for a period of 7 days followed by air-curing at room temperature. 2.2 Permeability Measurement

A test setup was developed to measure permeability according to the test scheme recently proposed by Li and Chau [26]. The test set-up essentially consists of a water-tight cell that houses another cell containing the hollow concrete specimen. The specimen is watertightly sealed at top and bottom using rubber gaskets. It is then subjected to a constant water pressure and the steady-state flow of water through the thickness of the hollow specimen is measured for calculation of permeability using the Darcy's formula. 2.3 Tortuosity Measurement

The tortuosity () of the each concrete specimen was calculated from the experimental data obtained from gas diffusion test and the pore properties of concrete (i.e. porosity, , and mean pore radius, rm). The calculation for tortuosity is based on the procedure as suggested by Navez [27]. Gas diffusion tests were carried out on the concrete specimens using the test set-up developed by Sharif et al. [23]. The set-up consists of a single pallet cell, gas flow lines, a detector, and a recorder. The concrete specimen used for the gas diffusion test was a circular disc of 2 inch diameter and 3/8 inch thickness. The porosity tests were conducted on 1 inch diameter core drilled out from the concrete specimens using a Vertical Automatic Drill Press. Normal water was used as a drilling fluid.

Then the core edges were trimmed using a Diamond Saw Blades Cutter. The finished cores were dried at 100C in a vacuum oven for 3 hours. The length, diameter and dry weight of each sample were measured. The grain volume of each sample was determined using PHI220 Helium Porosimeter. Sample porosity () was determined by dividing pore volume of the sample by its bulk volume. Mercury Intrusion Porosimetry (MIP) tests were conducted using Autoscan-33 Porosimeter to determine the mean pore radius (rm). The Autoscan-33 Porosimeter is designed to measure pore volume to a minimum radius of 0.0032 micrometers (32 Angstrom). This equipment consists of a sample cell assembly, filling apparatus, and a high pressure cell assembly. A cored and trimmed-to-size (around 10 mm diameter) sample of approximately two grams is assembled in the sample cell for low-pressure measurements. This sample cell is inserted in the vacuum jar of the filling apparatus and evacuated to at least 50 micron of mercury pressure. Then the test is performed applying pressure of up to 24 psi (0.165 MPa) to the sample cell and the volume of mercury intruded is measured simultaneously. The sample is then transferred to the high pressure cavity assembly where a gradually increasing capillary pressure is applied in the range of 0 to 33,000 psi (228 MPa) while measuring the volume of mercury that intruded in the sample. Withdrawal by reducing pressure to atmospheric pressure (14.7 psi (0.10 MPa)) and a re-injection by raising pressure again to 33,000 psi (228 MPa) are applied. High resolution pressure versus mercury saturation volume data are recorded for interpretation with the help of a computer linked to the system. Data from the low-pressure and high pressure assemblies are combined and processed for plotting the capillary pressure curves and interpreting the pore-throat-size distributions. Using the porethroat-size distribution data, the histogram of the pore-size distribution is plotted. The representative pore radius of a sample is determined from a log normal distribution function of the cumulative pore size frequency in percentage. The mean radius corresponds to 50% cumulative frequency. 2. Results and Discussion All the test results obtained for the 9 concrete mixes are presented in Table 2. Table 2: Test results Mix No. 1 2 3 4 5 6 7 8 9 CA/FA ratio (by wt.) 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 w/c ratio (by wt.) 0.4 0.4 0.4 0.5 0.5 0.5 0.6 0.6 0.6 Cement content (kg/m3) 300 350 400 300 350 400 300 350 400 Permeability coefficient, K (m/s 1012) 4.8 4.5 4.1 7.5 5.6 5.2 47.2 44.6 40.1 Porosity (%) 9.58 11.08 11.27 10.22 10.30 7.84 11.14 10.49 14.32 Mean Pore radius , rm() 829 821 728 652 600 889 614 907 800 Tortuosit y 200 245 275 116 99 119 38 61 78

Plots of water/cement ratio versus permeability coefficient, K, for different cement contents have been shown in Figure 1. It is clearly observed from Figure 1 that permeability coefficient increases in value with increase in w/c ratio. A relatively much sharper rise in permeability is noted when w/c ratio approaches 0.6. Results also show that cement content has some influence on permeability. Higher cement content at constant w/c reduces permeability to some extent. From the plots of Figure 1, it would appear that K-values are relatively insensitive to mix designs used in this study with w/c ratio from 0.40.5. Plots of water/cement ratio versus tortuosity,, for different cement contents have been shown in Figure 2. The values of tortuosity, plotted in Figure 2, show that tortuosity decreases with higher w/c

ratio in all cases of cement content. As the w/c ratio increases from 0.4 to 0.6, a sharp decrease is noted for mixes with different cement contents.

Cement content = 300 kg/cu.m

Cement content = 350 kg/cu.m

Cement content = 400 kg/cu.m

50 Permeability (x 1E-12 m/s) 45 40 35 30 25 20 15 10 5 0 0.3 0.4 0.5 water/cement ratio 0.6 0.7

Figure 1: Variation of permeability with mix parameters.

Cement content = 300 kg/cu.m

Cement content = 350 kg/cu.m

Cement content = 400 kg/cu.m

300 250 200 150 100 50 0 0.3 0.4 0.5

water/cement ratio

Tortuosity

0.6

0.7

Figure 2: Variation of tortuosity with mix parameters.

The experimentally determined values of permeability coefficient K and tortuosity for all 9 mixes are shown in Figure 3 to observe the relationship between the two properties. As seen from Figure 3, permeability has an inverse nonlinear relationship with for less than 125. For greater than 150, permeability is marginally affected by tortuosity.

60 Permeability (x E-12 m/s) 50 40 30 20 10 0 0 50 100 150 Tortuosity 200 250 300

Figure 3: Variation of permeability with tortuosity. 3. Conclusions Based on the observations made through this study, the following conclusions may be drawn: 1. Water/cement ratio has a significant impact on permeability and tortuosity. Permeability increases with increase in water/cement ratio, conforming to the general perception, whereas tortuosity decreases with increasing water/cement ratio. Permeability increase more rapidly with water/cement ratio approaching or exceeding 0.6 2. Higher cement content at a constant w/c ratio appears to reduce permeability and increase tortuosity to some extent. 3. The study shows that permeability is inversely related to tortuosity for values less than 150 or so. For exceeding 150, K-values are marginally affected by . Thus, for a low-permeable concrete, values are expected to be higher than 150. Acknowledgement The authors acknowledge the support provided by King Fahd University of Petroleum & Minerals, Dhahran, Saudi Arabia, under the ARI research grant (Project No. ARI-016). References [1] Collins, J.F., Derucher, Jr., K.N., and Korfiaris, G.P. Permeability of Concrete Mixtures Part 1: Literature Review, Civil Engineering for Practicing and Design Engineers, v. 5, 1986, pp. 579-638. Mehta, P.K. and Monteiro, P.J.M. Concrete: Structure, Properties and Materials, 2nd Edition, Prentice Hall, Inc., New Jersey, 1993.

[2]

[3]

Halamickova, P. and Detwiler, R.J. Water Permeability and Chloride Ion Diffusion in Portland Cement Mortars: Relationship to Sand Content and Critical Pore Diameter, Cement and Concrete Research, v. 25, n. 4, 1995, pp. 790-802. Nayme, B.K. and Illston, J.M. Capillary Pore Structure and Permeability of Hardened Cement Paste, 7th International Congress on Chemistry of Cement, Paris, v. III, Chapter VI, 1980, pp. 181-186. Mehta, P.K. and Manmohan, D. Pore Size Distribution and Permeability of Hardened Cement Pastes, Proceedings of the 7th International Conference on Chemistry of Cements, Paris, v. 3, 1980. Hughes, D.C. Pore Structure and Permeability of Hardened Cement Paste, Magazine of Concrete Research, v. 37, n. 133, 1985, pp. 227-233. Banthia, N. Water Permeability of Cement Paste, Cement and Concrete Research, v. 19, 1989, p. 727. Haynes, H.H. Permeability of Concrete in Sea Water, American Concrete Institute SP-65, Detroit, v. 1, 1980, pp. 21-38. Oberholster, R.E. Pore Structure, Permeability and Diffusivity of Hardened Cement Paste and Concrete in relation to Durability: Status and Prospects, Proceedings of the 8th International Congress on Chemistry of Cement, Rio de Janeiro, Sub-Theme 4.1, 1986, pp. 323-335. Powers, T.C., H.M. Mann, and Copeland L.E. The Flow of Water in Hardened Portland Cement Paste, Special Report No. 40, Highway Research Board, Washington, 1958, pp. 308323. Cui, L. and Cahyadi, J.H. Permeability and Pore Structure of OPC Paste, Cement and Concrete Research, v. 31, 2001, pp. 277-282. Kumar, R. and Bhattacharjee, B. Assessment of Permeation Quality of Concrete through Mercury Intrusion Porosimetry, Cement and Concrete Research, 34, 2004, pp. 321-328. Karger, J. and Ruthven, D.M. Diffusion in Zeolites and Other Microporous Solids, John Wiley and Sons, New York, 1986. Dullien, F.A.L. Porous Media, Fluid Transport and Pore Structure, Academic Press, New York, 1979. Bear, J. Dynamics of Fluids in Porous Media, American Elsevier Publishing Co., New York, 1972. Harak, Z. and Schneider, P. Chemical Engineering Journal, v. 2, 1971, p.26. Amberg, C.H. and Echigoya, E. Canadian Journal of Chemical Engineering, v. 39, 1961, pp. 215. Satterfield, C.N. Mass Transfer in Heterogeneous Catalysis, MIT Press, Cambridge, 1970. Petersen, E.E. American Institute of Chemical Engineering Journal, v. 4, 1958, p. 343. Michaels, A.S. American Institute of Chemical Engineering Journal, v. 5, 1959, p. 270. Bhatia, S.K. Journal of Catalysis, v. 93, 1985, p. 197. Bhatia, S.K. Chemical Engineering Journal, v. 41, 1986, p.1311.

[4]

[5]

[6] [7] [8] [9]

[10]

[11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22]

[23]

Sharif, A., Loughlin, K.F., Azad, A.K., and Navez, C.M. Determination of Effective Chloride Diffusion Coefficient in Concrete via a Gas Diffusion Technique, ACI Materials Journal, American Concrete Institute, v. 94, n. 3, May-June 1997, pp. 227-233. Gudmundsson, T. An Approach to Determining the Water Vapour Transport Properties of Building Materials, Nordic Journal of Building Physics, v. 3, 2003, pp. 1-10. Poupeleer, A.S., Carmeliet1, J., Roels, S., and Van Gemert, D. Validation of the Salt Diffusion Coefficient in Porous Materials, International Journal for Restoration of Buildings and Monuments, v. 9, n. 6, 2003, pp. 663-682. Li, Zong Yin and Chau, Chung-Kong. New Water Permeability Test Scheme for Concrete, ACI Materials Journal, v. 97, n. 1, January-February 2000, pp. 84-90. Navez, C.M. Chloride Diffusion in Concrete and Prediction of the Onset of Corrosion in Reinforced Concrete Structures, M.S. Thesis, King Fahd University of Petroleum & Minerals, Dhahran, 1994.

[24] [25]

[26] [27]