SOFTWARE-ENHANCED METHOD FOR RAPID DETERMINATION OF THE EARLY HEAT OF HYDRATION OF CEMENT CEM II/A AND B S TO PREDICT THE 28-DAY COMPRESSIVE STRENGTH

Transcription

1 The First International Proficiency Testing Conference Sinaia, România 11 th 13 th October, 07 SOFTWARE-ENHANCED METHOD FOR RAPID DETERMINATION OF THE EARLY HEAT OF HYDRATION OF CEMENT CEM II/A AND B S TO PREDICT THE 28-DAY COMPRESSIVE STRENGTH Maria Ioan 1, Liliana Radu 1, Constanta Mandoiu 2 1 CEPROCIM SA, 6 bd Preciziei, Code: 0623, Bucharest, Romania; 2 CARPATCEMENT HOLDING SA, Fieni, Romania 1 analize3@ceprocim.ro, 2 constanta.mandoiu@carpatcement.ro Abstract The paper dealt with prediction of the of cement by a rapid method of determining the early heat of hydration. The early heat of hydration and the standard strength of cement being determined on a number of cements may serve for establishing a correlation to predict the 28-day strength. The relation needs validation on a greater number of cements of the same type and manufacturer. The investigations involved the following steps: i) determining the standard strength as per SR EN 196 1, ii) determining the early heat of hydration by a calorimeter enhanced with 1-minute temperature increase monitoring software, iii) establishing a correlation equation, and iv) validating the correlation on 70 cements CEM II/A and B S. The rapid estimation of the 28-day strength of cement CEM II/A and B S with a precision of ±2.5 MPa is a versatile tool in the cement manufacturing process. The same algorithm may be applied and validated for any cement sort. Key words Cement, early heat of hydration, rapid method, prediction, standard strength 1 INTRODUCTION The paper aimed at establishing correlations between the heat of hydration and the strength of cement so as to enhance quality parameters and management in cement plants and streamline information transfer from testing laboratories to end-users. Solacolu [1] suggested the need of a rapid assessment of the cement strength by processing experimental information using charts of equal strength and modular 355

2 coordinates, keeping free lime constant, and highlighting the effect of SM and AM on the standard strength. Later, the effect of free lime CaO [2] as well as tricalcium aluminate C 3 A [3, 4] was taken into consideration. The effect of cement fineness and the contribution of various particle size fractions on short- and long-term strength development was studied as well [5 8]. Moreover, experimental methods were devised for rapid cement strength assessment. To this end, de Squeira Tango [9] presented a method for predicting the compressive strength of hydraulic cement on the basis of the 2- and 7-day compressive strength; Izaguirre [10] presented a method of determining the strength by curing at temperatures exceeding 100 C; and Ioan et al. [11] presented a method of predicting the strength by using the 1-day compressive strength as determined on thermally cured mortar specimens. Although literature failed to indicate any direct correlation between the strength and the heat of hydration of cement, studies carried out in the CEPROCIM testing laboratory on 30 cements CEM I from the same manufacturer revealed a linear correlation y = ax + b, with y being the determined as per standard SR EN [12] and x being the early heat of hydration released before reaching the maximum temperature θ M. Preliminary investigations on cement mortars were continued on cement pastes based on cement CEM I. Moreover, an accelerator was used to perform a rapid determination [13] of early heat of hydration. The present investigation focused on determining the heat of hydration by using a semi-adiabatic method on cement pastes based on cement CEM II/A, B-S. 2 EXPERIMENTAL 2.1 Materials a) Cements The laboratory experiments used 70 samples (S) of cement with addition of blast furnace slag (BFS) type CEM II from CARPATCEMENT HOLDING SA, Fieni. Figure 1 shows the oxide composition, the BFS content, and the 28-day compressive strength determined as per SR EN b) Sand The sand employed in the investigations was standard sand as per CEN EN (France) for determining the. c) Accelerator An accelerator was employed to determine the early heat of hydration Q hm released up to attaining the maximum temperature θ M. d) Water Distilled water as per SR EN was employed for reference testing in determining both the strength and the early heat of hydration. 356

3 65 Oxides, BFS (%); Strenght (MPa) CaO SiO 2 Al 2 O 3 Fe 2 O 3 MgO SO 3 BFS Strenght Figure1 Characterisation of the cements 2.2 Working technique Experimental technique A calorimeter prototype enhanced with software for 1-minute temperature increment monitoring with a precision of ±0.01 C was employed to calculate the early heat of hydration (Fig. 2). d b a c f e g Imagine de ansamblu a prototipului realizat (vedere din plan superior) Imagine a prototipului pregatit pentru începerea încercarii Figure 2 Calorimeter a temperature module; b temperature controller; c PELTIER cooler; d fan; e reference calorimeter; f calorimeter; g vacuum container Calculation technique Establishing a correlation between the standard strength and the heat of hydration involved two steps: I and II [13]. Step I: Automatic calculation of the early heat of hydration Q hm of cement up to attaining the maximum temperature in the calorimeter by using equation (1) (Figure 3). 357

4 Heat of hydration, Temperature Early heat of hydration (J/g) Maximum temperature ( 0 C) Figure 3 Heat of hydration calculated up to attaining the maximum temperature The early heat of hydration was calculated at 1-minute intervals by means of a software application program attached to the calorimeter. The heat of hydration Q (J/g) was calculated by adding the heat A contained in the calorimeter and the heat B lost in the environment. Q = A + B (1) The heat A was calculated using the overall heat capacity c of the calorimeter, the cement weight m c and the temperature increase θ t of the cement specimen at the time t. c A = θ t m c (2) The overall heat capacity c of the calorimeter including the cement specimen box and the cement specimen was calculated using eq. (3). c = 0.8 m c m w m acc m b + μ (3) Where: 0.8 specific heat capacity of cement, J/K.g 3.8 specific heat capacity of water, J/K.g specific heat capacity of aluminium, J/K.g specific heat capacity of CaCl 2, J/K.g μ default heat capacity of the empty calorimeter, J/K m c weight of cement, g m acc weight of CaCl 2, g m w weight of water, g m b weight of the cement specimen box, g. The heat loss was calculated over periods of hydration such as between successive iterations of temperature determination. The calorimeter heat loss coefficient α, expressed in J/h.K, was given by equation (4). α = a + b θ (4) Where: a, b calorimeter default constants θ temperature increase, K. 358

5 Step II: Establishing a correlation y = ax + b to predict the 28-day compressive strength y as a function of the early heat of hydration x released up to θ M. This step involved Graphical representation of the (standard strength) as determined experimentally as a function of the rapidly-determined early heat of hydration released up to attaining the maximum temperature (Fig. 4) Plotting the regression curve using equation y = ax + b (Fig. 4) Calculating the by using the obtained equation, and Calculating the differences between the experimental data and the calculated values. 28day compressive strength, MPa y = *x R = Early heat of hydration, J/g Figure 4 - Standard strength vs. rapidly-determined early heat of hydration Figure 4 presents the correlation between the 28-day strength determined on the 70 samples as per SR EN and the early heat of hydration determined as per the rapid method. The early heat of hydration is used for predicting the 28-day compressive strength by means of equation (5). y = 0,29471 x 13,05315 (5) 3 RESULTS 3.1 Effect of blast furnace slag addition Another objective of the experiments was to monitor the effect of the BFS addition on the, the early heat of hydration and the maximum temperature during cement hydration. The data regarding the 28-day compressive strength, the early heat of hydration and the maximum temperature during cement hydration as a function of the BFS content were presented in figures 5, 6 and

6 Slag (%), Rc28 (MPa) Slag (%) Rc28 (MPa) Figure 5 - Variation of the (red) and BFS content (black) for the 70 samples Slag (%), QhM (J/g) Slag QhM Figure 6 - Variation of the early heat of hydration (red) and BFS content (black) for the 70 samples 70 Maximum temperature ( o C) 30 Slag (%) Max. temp ( o C) Figure 7 - Variation of BFS content (black) and maximum temperature θ M (red) in the calorimeter for the 70 samples 3

7 3.2 Compatibility of calculated strength with experimental data Data in Figure 3 revealed that the correspondence relation between the 28-day compressive strength values determined experimentally and calculated using the established correlation y = x was of R = Figure 8 presents the strength values calculated on the basis of the above equation along with the experimental data. Differences between experimental data and calculated data varied within ±2.5 MPa. Exp Calc 28d strength, MPa 30 Max increases: 2.5 MPa Max decreases: 2.5MPa Number of samples Figure 8 - determined experimentally and calculated using the equation y = x (x = Q hm ) 4 CONCLUSIONS The paper dealt with establishing a correlation relation y = ax + b between the 28-day compressive strength y and the rapidly-determined early heat of hydration x of cement by employing a number of 70 samples of cement with addition of blast furnace slag CEM II/A, B-S from CARPATCEMENT HOLDING SA, Fieni. The obtained results revealed differences of ±2.5 MPa between the experimental data and the calculated values. The blast furnace slag addition was found to influence in the same manner both the early heat of hydration and the standard strength. A correlation equation for predicting the standard strength on the basis of the rapidlydetermined early heat of hydration is specific to each cement manufacturer and type and sort of cement. The method is new, reproducible and fast, with the time to predict the standard strength for cement CEM II/A and B-S being of 3-4 hours upon manufacture. It may be employed by any cement manufacturer both when changing the formulation and the cement sort. 361

8 REFERENCES [1] Solacolu, S. : La chimie physique des silicates techniques (in Romanian), 2 nd Edition, Editura Tehnică, Bucharest, 1968 [2] Schmitt-Henko, C. : Effect of clinker composition on setting and early strength of cement, ZKG International, 26 (2), 63, (1973) [3] Sylla, H.M. :Process Technology of Cement Manufacturing, Bauverlag GmbH, Wiesbaden, (1993) [4] Shalam, M. Ish; Bentur, A. :Effects of aluminate and sulfate contents on the hydration and strength of Portland cement pastes and mortars, Cem. Concr. Res., 2 (6), 653, (1972) [5] Locher F.W. et al : Influence of the fineness of grinding and the particle size distribution on the properties of Portland and blast furnace cements and hydraulic limes, Tonind-Ztg, 90 (12), 547, (1966) [6] Locher, F.W. :Setting and early strength of cement, ZKG International, 26 (2), 53, (1973) [7] Teoreanu, I. et al., Concrete Durability (in Romanian), Editura Tehnică, Bucharest, 1982 [8] Locher, F.W. et al : The effect of the particle size distribution on the strength of Portland cement, ZKG International, 26 (8), 349, (1973) [9] Squeira Tango, C.E. :An extrapolation method for compressive strength prediction of hydraulic cement products, Cem. Concr. Res., 28 (7), 969, (1998) [10] Izaguirre, J.R. :Procedimiento rapido para identificar las resistencias de los cementos, Cemento Hormigon, 65 (), 754, (1995) [11] Ioan, M. et al : A rapid method of determination of the 28-day compressive strength of cement on thermally treated specimens; CEPROCIM Technical Report, 00 [12] ***, SR EN 196-1:06, Methods of testing cement. Part 1: Determination of strength [13] Ioan, M. :CALIST Research Programme, Project No.4217 (03 05) [14] ***, SR EN 196-9:04, Methods of testing cement. Part 9: Heat of hydration. Semi adiabatic method 362