Construction and Building Materials 31 (2012) 341–346

Contents lists available at SciVerse ScienceDirect

Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat

Prediction of standard compressive strength of cement by the electrical resistivity measurement Xiaosheng Wei a,⇑, Lianzhen Xiao b, Zongjin Li c a

School of Civil Engineering and Mechanics, Huazhong University of Science and Technology, Wuhan 430074, China School of Materials Science and Engineering, Wuhan Institute of Technology, Wuhan 430073, China c Department of Civil Engineering, The Hong Kong University of Science and Technology, Hong Kong, China b

a r t i c l e

i n f o

Article history: Received 30 August 2011 Received in revised form 17 December 2011 Accepted 23 December 2011 Available online 4 February 2012 Keywords: Standard compressive strength Cement Hydration Resistivity Microstructure

a b s t r a c t This study was undertaken to suggest a rapid testing method which would permit prediction of standard compressive strength of a cement within 24 h. Eight cement pastes were prepared with three strength grades of cements at a constant water–cement ratio of 0.4 for electrical resistivity measurement during first 24 h and eight mortars were prepared with the same cements as those used in the pastes according to the procedures in ISO679:1989 for standard compressive strength tests of cement at 28 d. The results from electrical resistivity measurement show that the higher electrical resistivity value of the pastes is reached when the higher strength cement is used. A linear relation is established between the standard compressive strength of cements at 28 d and the resistivity of the pastes at 24 h. It is possible that an accelerated and non-destructive method could be developed to estimate the standard compressive strength of cement based on a quantitative relationship with electrical resistivity at 24 h. Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction The cement factories in China produce and supply cement by different strength grades. There are different strength grades of cement used in concrete mixtures, such as 32.5, 42.5 and 52.5 for various strength concrete mixtures. The strength grades 32.5, 42.5 and 52.5 mean that the standard compressive strength values of cement at 28 d are not lower than 32.5 MPa, 42.5 MPa and 52.5 MPa, respectively, by the test procedures in standard ISO 679 [1] and strength values 32.5 MPa, 42.5 MPa and 52.5 MPa are taken as nominal strength of a cement to be used in concrete. It is necessary to obtain cement actual strength at the age of 28 days (ISO679) as an important input parameter when a concrete mix proportion is designed. Usually, the cement strength is determined using its nominal strength and an empirical coefficient. In this way, cement is wasted because the concrete is too strong, or the concrete does not gain the intended strength due to variable cement strength value of a cement, even though cement comes from the same factory supplied by the same strength grade. Under the currently quicker pace of construction, the rapid determination of the actual cement strength is desired to be developed. Tsivilis and Parrisakis [2] used a mathematical model to predict cement strength and Relis and Soroka [3] used an accelerated testing for obtaining standard cement strength. Additionally, rapid methods ⇑ Corresponding author. Tel.: +86 27 87393312. E-mail address: [email protected] (X. Wei). 0950-0618/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.conbuildmat.2011.12.111

are desired in determination or prediction of concrete strength. Lapinas [4] used a modified boiling method for concrete strength testing, and Akkaya et al. [5] used an ultrasonic wave technique to predict the concrete strength at early ages. Rui Miguel Ferreir and Said Jalali [6] proposed an alternative method to estimate 28 d compressive strength of concrete based on the 7 d electrical resistivity measurements. The relationship of compressive strength of a cementitious material and its initial water–cementitious material is given by Bolomey formula [7,8]. Bolomey formula is widely accepted and used to calculate concrete strength at 28 d during concrete mix design, which is a linear relation between the ordinary concrete compressive strength and cementitious material–water ratio (C/W) as given in Eq. (1).

 fc;28 ¼ Afce

C B W

 ð1Þ

where fc,28 and fce are cube concrete compressive strength (MPa) at 28 d, cement nominal compressive strength (MPa) at 28 d; A and B are empirical constants related to aggregate type used. From Bolomey equation, it can be seen that an ordinary concrete strength is determined by cement strength, type of aggregate and initial water–cementitious material ratio (W/C). Initial W/C ratio of a concrete mainly influence final strength of concrete when a cement and a type of aggregate are chosen. From Power’s law [9] as given in Eq. (2), gel/space is determined by hydration degree and initial W/C. Concrete strength is correlated to gel/space and the higher

342

X. Wei et al. / Construction and Building Materials 31 (2012) 341–346

gel/space means the more hydration products and the higher strength of concrete,

Electrical resistivity measurment Transformer

gel 0:68a ¼ space W=C þ 0:32a

ð2Þ

Cement strength (ISO679)

I

W/C=0.5 C:S=1:3 T=20± 2° C

W/C=0.4 T=20± 2° C

V

Paste in a ring mold

where a is hydration degree of cement. A relationship between electrical resistivity of cement pastes and hydration degree, initial W/C has been obtained as given in Eq. (3) [10],



W=C DC  DH qðtÞ ¼ q0 ðtÞ  V total =   aðtÞ DW DC  DH

ρ ~ t from electrical resistivity response ρ24h

mðtÞ

f ~ t from compressive test f28d

ð3Þ t24h

where q(t) is the bulk electrical resistivity; q0(t) the pore solution electrical resistivity; u(t) the porosity (full of the pore solution); m(t) the cementation factor; Vtotal the total volume of each constituent in the concrete; Dw the density of the pore solution; Dc the density of cement; Dh the density of hydrates; and a(t) is the hydration degree of cement. It can be seen from Eq. (3) that the electrical resistivity is mainly dominated by the initial W/C ratio and the hydration degree or porosity during hardening period in a cement hydration system [11–13], and the electrical resistivity development of a cement paste is a fingerprint of increase in the degree of hydration, reversely decrease in porosity. The objective of this research is to investigate the correlation between the standard compressive strength of cement and electrical resistivity by monitoring the ISO679 compressive strength of cement and the electrical resistivity of the cement pastes with use of various strength grade cements. It is hard to vibrate the mortar mixes made by the proportion from standard ISO 679 in the mold of the resistivity setup, therefore, paste samples were used in electrical resistivity measurement and the electrical resistivity of pastes is introduced in this study as the index of microstructure formation and strength development, due to the properties characteristics evolution with time resulting from the hydration of the cement part. The SEM photos at different ages are used to understand the correlation between resistivity and compressive strength from microstructure scale on hydrated cement pastes. This study aims to develop an accelerated testing method to determine the standard compressive strength of cement. The flowchart of the determination of the relationship for the resistivity response and cement strength is shown in Fig. 1.

2. Materials and sample preparation Eight cements from different factories with three strength grades 32.5, 42.5 and 52.5 were used in eight cement pastes and eight mortar samples, respectively, which were commercially available in China. The cements meet the requirements of Chinese cement standard GB/T17671-1999, and the physical and mechanical properties are shown in Table 1. The mix proportions of eight cement pastes for electrical resistivity testing were shown in Table 2. The cements with strength grade 32.5 from four factories (marked as PO32.5-1, PO32.5-2, PS32.5-3 and PO32.5-4) were used in Mixes 1– 4, respectively. Mixes 5–6 used cements with strength grade 42.5 from two factories (marked as PO42.5-5 and PO42.5-6), respectively. Mixes 7–8 used cements with strength grade 52.5 from two factories (marked as PO52.5-7 and PO52.58), respectively. A constant W/C of 0.4 by weight was used in Mixes 1–8. Each paste sample was mixed for 2 min in a planetary-type mixer at 45 rpm and for a further 2 min at 90 rpm and cast in a ring-shape mold to measure electrical resistivity during 24 h. Eight mortar mixes for measuring compressive strength were prepared with use of the same cements as those used in the electrical resistivity testing. The mortar proportions and mixing procedures follow standard ISO679:1989.

t 28d

the relationship for resistivity response and compressive strength f 28d = a ρ24h + b Fig. 1. Determination of the relationship of resistivity and cement strength (ISO679).

3. Experimental methods 3.1. Electrical resistivity The electrical resistivity (ER) of the cement pastes was measured by a non-contact electrical resistivity apparatus, as shown in Fig. 2. The transformer principle was adopted in this apparatus and described in Ref. [14]. When an AC voltage from the generator was applied to the primary coil of the transformer, a toroidal voltage would be inducted in the ring sample that acted as the secondary coil of the transformer. The toroidal current could be measured by a leakage current meter and hence the resistivity of the sample could be obtained based on Ohm’s law. A calibration was done with 0.1 N KCl electrical conductivity at 24.5 °C using this device [12]. The difference between the measured value and the standard value is less than 0.4%, indicating that the setup provides high accuracy. After mixing, a fresh mixture was cast into a ring-shaped mold (1.67L) with a cover for preventing water evaporation during test period. Measurements were taken at intervals of 1 min for 24 h. The electrical resistivity was calculated by the actual height of the measured sample. The bulk electrical resistivity q(t) of a cement paste depends on the ionic concentration in the liquid phase and the volume fraction of liquid phase. Archie15 studied the relationship between the resistivity and the porosity of the rocks saturated with conducting water. A modified relationship from Archie’s law can be applied to cement paste as given in Eq. (4).

qðtÞ ¼ q0 ðtÞ  a/ðtÞm

ð4Þ

where parameters q(t), q0(t), u(t) and m have the same meaning as those in Eq. (3). 3.2. Compressive strength The cement strength was measured based on standard ISO679:1989 (GB/T 17671-1999). Three mortar mixes were prepared for determining standard strength of each cement in prisms

343

X. Wei et al. / Construction and Building Materials 31 (2012) 341–346 Table 1 The physical and mechanical properties of cements (ISO679:1989). Normal consistency (%)

Fineness (%)

PO32.5-1 PO32.5-2 PS32.5-3 PO32.5-4 PO42.5-5 PO42.5-6 PO52.5-7 PO52.5-8

32.0 27.6 26.4 26.2 28.2 27.0 27.4 27.0

2.84 4.76 8.56 5.36 0.56 0.48 0.88 0.00

Setting time (min)

Flexual strength (MPa)

Compressive strength (MPa)

Initial

Final

3 (d)

28 (d)

3 (d)

28 (d)

180 165 199 180 196 172 159 160

216 288 240 223 244 221 211 318

2.22 3.00 2.63 4.02 4.77 5.23 5.92 5.00

5.82 6.40 6.37 8.03 7.95 8.02 8.53 8.40

12.8 19.7 14.1 22.3 25.4 29.2 31.3 28.9

32.2 36.1 37.3 44.6 45.1 45.6 54.7 53.8

Table 2 The paste proportions for electrical resistivity tests and their electrical resistivity at 24 h. Mix no.

Cement grade with sample number

Water

Cement

q24 (Ohm m)

1 2 3 4 5 6 7 8

PO32.5-1 PO32.5-2 PS32.5-3 PO32.5-4 PO42.5-5 PO42.5-6 PO52.5-7 PO52.5-8

0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4

1 1 1 1 1 1 1 1

1.571 1.656 1.676 2.876 2.938 3.036 3.743 3.743

Electrical resistivity (Ohm.m)

Cement grade with sample number

4

3

2

8 7

Group 3

6 5 4

Group 2

3 2 1

Group 1

1

0

4

8

12

16

20

24

time (h) Fig. 3. The bulk resistivity developments of the eight pastes with W/C = 0.4 during 24 h.

Transformer Current meter Input/output cables

Paste sample in a ring mold

Fig. 2. The setup of the non-contacting resistivity measurement.

with size of 40  40  160 mm. The mortars were mixed, cast, compacted, and placed in a moist storage room (20 ± 2 °C) for 24 h. The mortar samples were demolded and moved to a standard curing room up to 28 d. The flexural tests of three prisms were firstly conducted and then the compressive strength tests of six samples obtained from the flexural tests for each mix were carried out. 3.3. SEM observations The hydrated pastes from Mix 8 for SEM observations were prepared by stopping hydration using acetone at ages of 0 h, 7 h and 24 h and the samples were then dried in a vacuum oven at room temperature for 24 h. 4. Experimental results and discussion 4.1. Bulk resistivity development during 24 h The bulk electrical resistivity developments with hydration time up to 24 h for the eight paste mixes are plotted in Fig. 3. The electrical resistivity values at 24 h (q24) were shown in Table 2. It can be found in Fig. 3 that all the curves of the eight samples follow a similar trend with time. The electrical resistivity q(t) drops to a minimum point, and then gradually increases with time. The decrease in resistivity right after mixing is due to immediate

dissolution of soluble ions (such as Ca2+, K+, Na+, SO2 4 ) from the cement particles after water is added to cement and the dissolving process of the ions causes the resistivity decrease in the very early period. Subsequently, there is a slight increase on the resistivity curves due to the formed hydration products blocking ions conductive paths after a certain supersaturation of the solution is reached. The following sharper increase in electrical resistivity is caused by a large amount of hydration products formation in the solid phase and finally a relative stable increase trend is reached by the ions diffusion control of hydration process. It can also be found in Fig. 3 that the electrical resistivity curves in the late period obviously fall in three groups, as group 1 (Mixes 1–3) having the lowest resistivity, group 2 (Mixes 4–6) in the middle and group 3 (Mixes 7–8) having the highest resistivity. In groups 1, 2 and 3, the cements with grades 32.5, 42.5 and 52.5 were used, respectively, except that the q24 of Mix 4 with grade 32.5 cement falls in the middle group (Mixes 4–6) since cement PO32.5-4 used in Mix 4 has an over high actual strength with comparison of the other grade 32.5 cement. With comparison to Mixes 1–3 (grade 32.5 cement used), Mixes 5–6 (grade 42.5 cement used) or Mixes 7–8 (grade 52.5 cement used) have the higher increments in bulk electrical resistivity from the minimum point occurrence to 24 h. The steeper increase of the electrical resistivity in the samples with the higher strength cement is attributed to more rapid porosity decrease rate and more hydration products being formed within the samples. The porosity decreases with hydration time, resulting from the increase of the solid phase in the hydration systems, which favors both electrical resistivity increase and compressive strength development. The higher strength cement has a higher reaction rate than that of the lower strength cement normally due to lower strength cement with lots of low active mineral admixtures added when it was manufactured in a factory. As known, the low active additives (such as fly ash, limestone powders) are usually used in cement clinker to reduce cement cost for various purposes. The standard

344

X. Wei et al. / Construction and Building Materials 31 (2012) 341–346

0h (amplified by 2000)

7h (amplified by 2000)

7h (amplified by 20000)

24h (amplified by 20000)

Fig. 4. SEM pictures at 0 h, 7 h and 24 h for sample PO52.5-8.

compressive strength of cement is then decreased and the strength grade of cement is lowered. With comparison of higher strength grade cement, there are higher contents of low active additives used in the lower strength grade cement as mentioned, resulting in a decrease of reactive compositions for hydration products formation. As the hydration develops, less hydration products develop in a lower grade cement system than that develops in a higher grade cement system. This leads to more voids and lower tortuosity in the sample. Hence, a lower q value is also reached. The low active additives lead to a significant dilution effect of cement and reduce the amount of hydrates. Such effect increased with low active additives percentage, which is similar to the observation in fly ash replacement of cement hydration system in Ref. [10]. It can be concluded that the electrical resistivity and the compressive strength have consistent development trend for the pastes, the higher electrical resistivity (q24) of a mix means that the higher compressive strength can be gained. The correlation between electrical resistivity and compressive strength can be analyzed from microstructure scale by SEM photos and a model as shown in Figs. 4 and 5, respectively.

(a) right after mixing

(b) hydrated for some time

Fig. 5. The schematic presentation of the conduction paths in a cement paste.

From the SEM photo at 0 h, cement particles are separated by mixing water in the hydration system. With time, the solid particles start to connect together with increase of the hydration products as shown in the photos at 7 h (amplified by 2000 and 20,000), corresponding resistivity value of 1.232 Ohm m for sample PO52.5-8. The more and larger fibrous products can be seen in 24 h photo, corresponding resistivity value of 3.743 Ohm m, when the solid hydration products connected with each other. The SEM results present that both the compressive strength gain and the resistivity development contribute to more hydration products being formed with time. The compressive strength and electrical demonstrate the process of the hydration products increase and microstructure formation in different aspects. The microstructure development from flowing state (cement particles separated in the paste), plastic state (the cement particles with a layer of solid products start to connect each other) to solid state (the solid particles networking) with time was also schematically shown in the model in Fig. 5. From schematic model in Fig. 5, it can be seen that right after mixing, all the pores are connected and the cement particles are separated in fresh cement paste as shown in Fig. 5a. After curing certain time, the capillary pores, the remnants of the initially water filled space in a hydrated cement paste are the main conducting phase, and the hydration products block the pores conduction path and tortuosity increases as shown in Fig. 5b. The decrease of connectivity and the decrease of the amount of free water results in a sharp increase in the resistivity curve. Meanwhile, the porosity decreases due to the increase and accumulation of the hydration products that also resulted in the gain and increase in compressive strength. The testing and analysis results obtained show that both resistivity and compressive strength demonstrate the hydration characteristics of cement pastes from conduction paths and loading

X. Wei et al. / Construction and Building Materials 31 (2012) 341–346

345

capacity, respectively. The increase of the solid hydration products with time or the decrease of the porosity causes the compressive strength gain and also blocks the conduction path for the increase of electrical resistivity. The correlation between the electrical resistivity and the compressive strength of cement pastes can also be demonstrated by two numerical models. The first model is the relationship between electrical resistivity and porosity based on an empirical Archie equation [15] as shown in Eq. (4). The second model is an equation about strength and porosity [16], as given in Eq. (5).

higher actual strength cement used in Mixes 4–6, indicating that the electrical resistivity response of the cement pastes is directly related to the actual strength of cement when the mixes have a fixed proportion. Thus, the resistivity measurement monitors changes in the strength gain introduced by the different cements, with a high level of sensitivity. Based on the current testing, the compressive strength of the samples at 28 d f28 can be expressed by the electrical resistivity at 24 h q24 as shown in Eq. (6) and the correlation is shown in Fig. 6. A general expression can be obtained as given in Eq. (7).

f ¼ f0 ð1  /Þn

f28 ¼ 8:76  q24 þ 20:4; R2 ¼ 0:96

ð6Þ

f28 ¼ a  q24 þ b ðgeneral expressionÞ

ð7Þ

ð5Þ

where f is the compressive strength of sample with porosity /; f0 the compressive strength at u = 0; and n is the power coefficient. From Eq. (4) and Eq. (5), the correlation between the compressive strength and resistivity can be interpreted by the most important transition parameter porosity u, from the micro-scale to the mechanical properties of concrete. Porosity is of interest because it directly influences both mechanical and transport properties of cementitious materials. The properties of porous cement-based materials, such as mechanical strength and transport properties, depend primarily on the pore space and the morphology of the solid phase. The development of electrical resistivity with time reflects the decrease of pore space and the increase of solid space. Both resistivity and strength present hydration characteristics of cementitious materials in different ways. The development curves of resistivity and strength follow similar hyperbolic trend [17]. The analysis from microstructure and the numerical equations demonstrates that there is intrinsic correlation between resistivity and strength and a quantitative relation can be attempted to be plotted. 4.2. Correlation of the cement strength measured by ISO679 and the resistivity response occurring at 24 h

Compressive strength for mortar at 28 days (MPa)

On the basis of the analysis earlier, the test results of resistivity and compressive strength were analyzed and a quantitative correlation of the cement strength at 28 d measured by the ISO679:1989 procedures (shown in Table 1) with the resistivity values at 24 h are desired. It can be seen in Table 2 that the development trend with hydration time is the same in both the f28 and the q24. The changing trend in the resistivity response q24 is the same as that observed in the compressive strengths (f28) caused by the different grades cement. The highest resistivity q24 appears in Mixes 7–8 correspond to the highest cement strength used, and the higher q24 are reached in Mixes 4–6 than those in Mixes 1–3 due to the

60 55 50

5. Conclusions Based on the results of current investigations, the following conclusions are drawn: (1) Electrical resistivity developments with hydration time of the pastes with different cements, follow a similar pattern, first drop to a minimum point, and then gradually increase with time. The resistivity of the pastes with higher cement strength is higher than that of the paste with lower cement strength when water cement ratio is fixed. (2) Based on the microstructure analysis, the numerical equations, the correlation between resistivity and strength has been established. According to current testing results, a linear relationship can be obtained as f28 ¼ a  q24 þ b. An un-known cement standard strength at 28 d can then be determined by measuring its electrical resistivity within 24 h.

f 28d = 8. 7648 ρ24 + 20. 406 R² = 0.9634

Acknowledgments

45

The authors wish to express their gratitude and sincere appreciation to the National Natural Science Foundation of China (51178202) and the Doctoral Fund from Wuhan Institute of Technology for financing this Research Work.

40 35 30 1.5

where f28 is the standard compressive strength of a cement at 28 d, in MPa; q24 the bulk electrical resistivity of cement paste when W/C = 0.4 at 24 h, in ohm m; and a, b is the practical parameters, depending on W/C of samples in electrical resistivity testing. It can be concluded that the electrical resistivity curves q(t)t dynamically reflect the internal microstructure formation process and strength development in cementitious materials and the resistivity is a measure of its strength. According to current study, it is possible to determine the standard cement strength of a cement by electrical resistivity measurement. A wider range of properties of cement will be needed for obtaining more practical parameters a and b based on more tests in further study. This finding is an important step towards establishing the electrical resistivity method as an accelerated and non-destructive method for estimating strength of concrete. It can be used to adjust mix proportion of concrete in construction and also correct mix proportion of raw materials in cement factories in time.

2.0

2.5

3.0

3.5

4.0

Electrical resistivity for paste at 24 h (ohm·m) Fig. 6. The relationship between compressive strength and electrical resistivity for the pastes.

References [1] ISO (International Standard Organization). ISO 679:1989 (GB/T 17671-1999). Methods of testing cements – determination of strength; 1989. [2] Tsivilis S, Parrisakis G. Mathematical model for the prediction of cement strength. Cem Concr Res 1995;25(1):9–14.

346

X. Wei et al. / Construction and Building Materials 31 (2012) 341–346

[3] Relis M, Soroka I. Prediction of standard compressive strength of cement by means of accelerated testing. Mater Struct 1874;7(2):103–9. [4] Lapinas RA. Accelerated concrete strength testing by modified boiling method: concrete producer’s view, SP 56. ACI Accelerated Strength Testing: American Concrete Institute; 1978. p. 75–93. [5] Akkaya Y, Voigt T, Subramaniam KV, Shah SP. Nondestructive measurement of concrete strength gain by an ultrasonic wave reflection method. Mater Struct 2002;36(8):507–14. [6] Rui Miguel Ferreir, Said Jalali. NDT measurements for the prediction of 28-day compressive strength. NDT&E Int 2010;43(2):55–61. [7] Bolomey J. Granulation et prevision de la resistance probable des betons. Travaux 1935;30(19):228–32. [8] Yeh I-Cheng. Generalization of strength versus water–cementitious ratio relationship to age. Cem Concr Res 2006;36(10):1865–73. [9] Powers TC, Brownyard TL. Studies of the physical properties of hardened Portland cement paste. Bulletin, Research Laboratories of the Portland Cement Association Chicago: vol. 22; 1948. [10] Lianzhen Xiao, Zongjin Li. Early-age hydration of fresh concrete monitored by non-contact electrical resistivity measurement. Cem Concr Res 2008;38(3):312–9.

[11] McCarter WJ, Chrisp TM, Starrs G, Blewett J. Characterization and monitoring of cement-based systems using intrinsic electrical property measurements. Cem Concr Res 2003;33(2):197–206. [12] Li Z, Wei X, Li W. Preliminary interpretation of Portland cement hydration process using resistivity measurements. ACI Mater J 2003;100(3):253–7. [13] Silva BJ, Jalali S, Ferreira RM. Estimating electrical resistivity based on early age measurements. In: Proceedings of the international RILEM workshop performance based evaluation and indicators for concrete durability, Madrid; 2006. [14] Xiao L, Wei X, Li Z. Concrete hyzdration based on electrical resistivity measurement. VDM Verlag Dr. Mueller e.K; 2008. [15] Archie GE. The electrical resistivity log as an aid in determining some reservoir characteristics. Trans Am Inst Min Metall Pet Eng 1942;146:54–62. [16] Odler I, Robler M. Investigation on the relationship between porosity, structure and strength of hydrated Portland cement paste. (II) Effect of pore structure and of degree of hydration. Cem Concr Res 1985;15(3):401–10. [17] Xiaosheng Wei, Lianzhen Xiao, Zongjin Li. Hyperbolic method to analyze the electrical resistivity curve of Portland cements with superplasticizer. J Wuhan Univ Technol (Mater Sci Ed) 2008;23(2):245–8.