Printer Friendly

Effect of the addition of sodium citrate on the properties of low cement refractory concrete.


It is well known that the properties of refractory concretes are governed by the filler used in their formulation. The binder component (normally high alumina cement) of the concretes imparts the strength required during transportation and erection; this strength is attained after setting and drying. During subsequent heating up to the temperatures preceding sintering, irreversible destructive processes occur, as a rule, as cement loses its water of hydration. Therefore, in order to improve the existing refractory concretes and to create new concretes, it is necessary to decrease the content of high-alumina cement in to the minimum possible extent or even to produce them without any cement binders (2, 3 4).

Thermally stable aggregates combined with high alumina cement are the principal ingredients of monolithic acidic refractories. These aggregates are available both naturally and artificially. Raw materials available in nature unavoidably vary slightly in their compositions. However it is important to take advantage of the characteristics of these natural minerals that cannot be developed artificially rather than to avoid their use due to variations of chemical composition.

Unlike natural raw materials, artificial raw materials allow adjustment of chemical composition as well as their mineral constituents, and it is possible to get a uniform quality.

One common type of aggregate is bauxite, a raw material for alumina containing about 60% alumina. When calcined (usually in rotary kilns), the alumina level usually exceeds 85%. Calcined bauxite contains corundum as its principal component, mullite and a small amount of glassy phase.

On the other hand, grog is an artificial aggregate usually obtained from crushed defective refractory bricks. Its alumina content depends on that of the original bricks. It usually ranges from 40 to 80%.

Other types of aggregate used in the production of acidic refractory concretes include diaspore ([Al.sub.2][O.sub.3]. [H.sub.2]O) and corundum ([Al.sub.2][O.sub.3]).

On the other hand, admixtures are used in the mixing and placing of refractory concretes to modify favorably the properties of the fresh and hardened concrete. They are frequently added to control working time and flow properties of the material. Water reduction by the use of suitable admixtures is extremely important as this can lead to a denser material with much improved properties. Typically an admixture may modify more than one property, e.g. reduce the water requirement and increase the working time for the castable. Typically used water reducing admixtures are sodium citrate, sodium silicate, sodium salt of CMC and phenolic resins (5,6,7,8). Water demand is reduced throughout the following mechanism: Water reducing agents are absorbed on the surface of the cement grains, increasing the zeta potential and promoting good dispersion due to the repulsion between positively charged grains. This explains the decrease in water of mixing following an increase in the level of sodium citrate added.

These investigations did not cover, however, the combined effect of varying cement content and additive level on the properties of unfired or fired concretes.

In the present paper are studied the physico--mechanical properties of refractory concrete samples prepared from bauxite and grog to which was added sodium citrate as water reducing agent.


Raw Materials

The raw materials used are:

* Refractory cement containing 50% alumina was obtained from Lafarge Cement.

* Calcined bauxite was obtained from the Alexandria Company for Refractories with an alumina content exceeding 80%.

* Grog was obtained from previously fired defective refractory bricks that were crushed, ground and screened.

* Sodium citrate was purchased from "Gomhouria" Company for Chemicals, Cairo, Egypt. Its purity was stated by the manufacturer to exceed 99.8%.

Particle size distribution of aggregates

Standard sieves were used to sort both types of aggregate used: bauxite and grog. This was made in accordance with ASTM D 422/2007 (9).

Scan Electron Microscopy

Scan Electron Microscopy (SEM) was used to follow the effect of adding sodium citrate on the porosity of prepared concrete samples. To this aim, a SEM apparatus (SEM; Model JSM-5410) was used.

Preparation of samples

Following the results previously obtained for optimum particle size distribution (1), three sized samples were prepared using aggregate having the following screen analysis:
% Weight

0-1 mm 1-3 mm 3-5 mm
40 45 15

The cement content of these samples varied from 10% to 20% with 2.5% step. The pastes were hand kneaded with an adequate amount of water, which was determined for each batch according to the standard "good ball in hand test (10)". To water was added sodium citrate in the following percentages: 0.5%, 1%, 1.5% and 2%. The mixed batches were then cast into cubes of 50 mm side length using a vibrating table for 4 minutes at a frequency of 50 Hz and 4 minutes. The cast samples were left in their moulds for 24 hours in a 100% relative humidity cabinet. The hydrated samples were then de-molded. The specimens were left in an open air until their moisture content reaches 3-6%, then put in the drying oven at (110 [+ or -] 5) [degrees]C until reaching constant weight. They were then tested for water absorption, bulk density and apparent porosity and cold crushing strength.

Apparent Porosity, Water Absorption, and Bulk Density

These properties are determined according to ASTM Standards C 20/2007 (11). For each test, the average measurements for five specimens at least are calculated.

The five specimens for each test are weighed to get their dry weight (D) for each. The test specimens are then placed in water and boiled for 2 h in a boiler and kept entirely covered with water with no contact with the heated bottom of the container. They are cooled to room temperature while still completely covered with water. The weight (S) of each test specimen is determined after boiling and while suspended in water. The saturated weight (W) is also determined.

Apparent porosity is calculated from:

P,% = [[W - D]/V] x 100 (1)

Water absorption is calculated from:

A,% = [[W - D]/D] x 100 (2)

While bulk density is calculated as follow:

[[rho].sub.B] = D/V x 100 (3)


P = apparent porosity, (%);

W = weight of the specimen as saturated with water, (g);

D = dry weight, (g);

S = weight of the specimen as suspended in water, (g);

V = exterior volume = W -S, ([cm.sup.3]);

A = water absorption, (%);

[[rho].sub.B] = bulk density, (g/[cm.sup.3]).

Cold Crushing Strength

This was done to determine the compression stress to failure of samples consisting of three specimens cured for 28 days. It represents an indication of its probable performance under load. Each specimen was placed between two plates of the compression strength tester. This was followed by the application of an axial uniform load. The load at which a crack appears on the sample was noted, and it is calculated according to BS EN Standards 993-5/2000 (12):

C.C.S([[sigma].sub.c]) = W/a (4)


[[sigma].sub.c] = cold crushing strength, (MPa);

W = total maximum load at 3% deformation or at visible failure, (N);

a = average of gross areas of the two faces, ([mm.sup.2]).

Fired properties

Samples fired at 1300[degrees]C for 1 h. were tested for cold crushing strength, spalling resistance and permanent linear change on reheating.

Spalling resistance was tested according to Egyptian Standards ES 6444-1/2007 (13). In this test, the average measurements for three specimens at least are calculated.

The three specimens for each test with dimensions (50x50x75) [mm.sup.3] are dried at (110 [+ or -] 5)[degrees]C. The test pieces are then placed in the cold furnace. The furnace is then heated at a uniform rate so that it attained the test temperature 1100[degrees]C in 3 hours. The testing temperature is maintained for 30 minutes and the test pieces then removed from the furnace with a pair of tongs. After being cooled for 10 minutes the test pieces are replaced in the furnace for a further 10 minutes and the cycle is repeated. At the end of each cooling period, the test pieces are examined for cracks or loss of corners. Specimens passing this test should not show any cracks or corner losses after 45 cycles.

The permanent linear change on reheating was determined according to ASTM C179/2009 (14) by reheating the fired specimens at 1200[degrees]C for 1 h. and measuring the dimensions before and after reheating. The procedure is repeated as long as the dimensions do not stabilize.

Results and Discussion

Effect of adding sodium citrate on water Consumption

Figs.1 and 2 show the amount of consumed water as function of the amount of sodium citrate used with both bauxite and grog as aggregates.



It is worth mentioning that the level of sodium citrate used was limited to 2% so as not to affect seriously the refractoriness of the prepared castables since the presence of sodium ions highly affects the refractoriness of concretes (15). It is clear from both figures that asymptotic values of water addition are reached and that it is sufficient for practical purpose to add 1.5% sodium citrate. For grog containing mixes, the percent water added in, say, the 20% cement case, drops from 10.6% (0% citrate) to 5.65% (1.5% citrate), a drop of about 47%, where as the further drop in adding 2% citrate amounts only to about less than 2%. A comparison between the two figures shows that the amount of mixing water is slightly lower in case of bauxite than grog containing mixes.

Also, when correlation analysis was performed on both types of products, it was apparent that the percent reduction in water of mixing is much more sensitive to variations in percent citrate addition than to the percent of cement used. Table 1 shows the correlation results for bauxite based samples.

Effect of adding sodium citrate on percent Water Absorption, apparent porosity and bulk density

Since these three properties are strongly related it was thought sufficient to show how the three of them were affected by additions. Figs. 3 and 4 show the effect of adding sodium citrate on the percent water absorption and apparent porosity of bauxite based samples. It is clear that both properties are equally affected by an increase in percent citrate added. For example, as the percentage citrate increases from 0 to 2%. the apparent porosity and water absorption for 20% cement samples drops from about 15.5 to 11% and from 5.5% to 3.3% respectively.



Figs. 5 and 6 show the effect of of adding sodium citrate on the percent of water absorption and apparent porosity when grog was used instead of bauxite as aggregate. The values of porosity and water absorption are slightly higher than in case of using bauxite although the trend remains the same: For example,, as the percentage citrate increases from 0 to 2%. the apparent porosity and water absorption for 20% cement samples drops from about 17 to 12% and from 6.1% to 3.7% respectively.



Fig. 7 shows the effect of adding sodium citrate on the bulk density of samples prepared by using either bauxite as aggregates. This figure shows that the bulk density increases for samples containing 20% cement from 2.8 to 3.4 g/cm as the percent citrate is raised from 0 to 2%. this is due to the reduction in porosity. Similar results were obtained for grog based samples although not shown. The increase in bulk density was slightly less than in case of bauxite based samples, presumably because of the higher intrinsic porosity of grog as compared to bauxite (1).


Effect of adding sodium citrate on cold crushing strength(CCS)

Cast cubes prepared by adding sodium citrate to the cement--aggregate--water mix were air cured for different periods of time and their cold crushing strength determined. Fig. 8 shows the effect of curing time on the average cold crushing strength of samples, for different citrate and for cement contents = 12.5% (as illustration) for mixes containing bauxite as aggregate. The as--cast strength (0 time) is very low regardless of the amount of citrate added. It ranges from about 5 to 7 MPa. As time elapses, the samples reach most of their strength in 2 days, subsequent curing for up to 12 days having little effect on further increase in strength. The early increase in cold crushing strength is more sensitive to curing time than to the amount of citrate added. The effect of this latter on strength is higher than that of the amount of cement added. When other percentages of cement were used, the trend was the same. Fig.9 shows the maximum values of strength achieved as function of both percent of citrate and that of cement.



A direct result from Fig.8 is that curing can be practically stopped after 4 days since prolonging curing will not affect the achieved strength. It is also apparent that higher values of strength can be obtained by increasing the level of citrate addition rather than increasing the cement dose which is an economic advantage owing to the continuous increase in the cost of the latter.

When grog was used instead of bauxite and the experimental steps repeated, the behavior of CCS with respect to variations in curing time, citrate content and cement content was similar to that in case of using bauxite as aggregate. The main difference is the relatively lower levels of strength compared to bauxite for reasons previously aforementioned as can be seen by comparing Fig.9 and Fig.10.


To outline the relative effect of the variation in cement content, percent addition and curing time on the cold crushing strength of samples, correlation tables were established. To avoid redundancy, only one such table is reported in what follows. Tables drawn for the other cases were very similar in nature.

It can be seen from Table (2) that curing time is by far the most important factor determining strength development. The effect of percent addition follows next by a large margin and cement content plays the least role.

Results of samples fired at 1300[degrees]C

Cold crushing strength

As previously stated, one of the drawbacks of using sodium acetate as water reducer is the well--known fact that the presence of sodium negatively affects the refractoriness of these samples. That is why; it was thought necessary to follow up the behavior of samples to which sodium citrate was added when fired to 1300[degrees]C. The most serious drop in properties should occur at the maximum addition of sodium citrate (2%). Figs.11 and 12 show the variation of compressive strength of fired samples containing bauxite and grog respectively compared with the maximum cold crushing strength of these samples.



These two figures show that the values of compressive strength of samples fired at 1300[degrees]C are generally higher than the values recorded at room temperature. This implies that the presence of sodium ions has not exerted a negative influence on the final strength. Also visual observation has not shown any glassy phase formation.

It is worth mentioning that sodium citrate contains 26% of its weight sodium. This means that on adding 2% sodium citrate, the actual percent sodium in the mix would be about 0.5% which is a permissible figure in the formulations of refractory materials. This is the probable reason why its presence did not affect the strength of fired bodies.

Spalling test

Two different samples fired to 1300[degrees]C were subjected to spalling test according to E.S. 6444-1/2007. These samples consisted of three specimens each: The first sample contained 2% citrate + 15% cement + Bauxite and the second contained 2% citrate + 15% cement + Bauxite. The specimens were subjected to 45 consecutive firing and air quenching cycles (For one hour at 1100[degrees]C). They were then checked for the appearance of any surface cracks.

Visual observation of all specimens after the whole cycle did not reveal any crack formation in any case.

Percent linear change on firing and reheating

The linear changes occurring on firing at 1300[degrees]C were determined for all samples by vernier caliber measurement. The shrinkage associated with any of these samples was barely noticeable: for example, grog samples containing 12.5% cement and 2% citrate showed a linear shrinkage of 0.36%, while bauxite samples showed no perceptible shrinkage whatsoever.

When fired samples were reheated according to ASTM C -179/2009 at 1200[degrees]C, there was no perceptible dilatation in any of these samples.


Sodium citrate was added in percentages up to 2% to refractory concrete mixes containing from 10 to 20% cement (50% alumina) and bauxite or grog as aggregate. The presence of sodium citrate helped decreasing water required for mixing thereby decreasing the porosity of the samples and increasing their cold crushing strength.

The level of sodium citrate used was limited to 2% so as not to affect seriously the refractoriness of the prepared castables. Asymptotic values of water addition were reached so that it was sufficient for practical purpose to add 1.5% sodium citrate. For grog containing mixes, the percent water added (on using 20% cement) dropped from 10.6% (0% citrate) to 5.65% (1.5% citrate), a drop of about 47%, where as the further drop on adding 2% citrate amounted only to about less than 2%. A comparison shows that the amount of mixing water was slightly higher for bauxite based formulations.

On adding sodium citrate, the results showed that for a 20% cement level, the apparent porosity decreases in case of using bauxite from 15.5% to 11.1%, while it decreases in case of using grog from 16.8% to 12%. This was made clearer by SEM micrographs.

It was also concluded that using as low as 1% citrate additive coupled with a low cement content may yield a strength comparable with that obtained on using 20% cement without additive: For example, adding 1% sodium citrate to a mix containing 10% cement + grog results in a 2 day strength of 40 MPa compared to only 35 MPa if 20% cement is used without additives.

Finally, it was found that for samples containing sodium citrate, the values of compressive strength of samples fired at 1300oC were generally higher than the values recorded at room temperature. This implied that the presence of sodium ions did not exert a negative influence on the final strength. Also visual observation has not shown any glassy phase formation. Such samples did not exhibit any cracks on testing them for spalling nor did they exhibit any linear change on reheating at 1200[degrees]C.


[1] Ghonaim, S.A., Ghazal, H.B.G., and Abadir, M.F., 2010,"Effect of Type of Aggregate on the Properties of Refractory Concrete". J. Am. Sc., 6(12), pp. 673-684.

[2] Otroj, S., Nilforoushan, M.R., and Marzban, R., 2009,"The Effect of Additives on the Properties of High Alumina Low-Cement Self-Flowing Castables", Ceramics, Department of Engineering., University of Shahrekord, Shahrekord, Iran Gadr (Niru) Refractories Co., Teheran, Iran, Silikaty 53 (1) pp. 42-47.

[3] Angelescu, N.G., Muthu, K.U., Ionita, G.G., Nicolae, A.N.A, and Nicolae, M.A., 2008,"Theoretical Fundaments and Applications about Special Additives", Valahia University Targoviste and Metallurgical Research Institute Bucharest, Romania, ICCBT 2008-A-(14), pp. 165-176.

[4] Belyakov, A.V., and Shayakhmetov, U.Sh., 1998,"Problems of Application of Unmolded Refractories", J. Mat. Sc., 33 (24), pp. 5827-5833.

[5] Lankard, D.R, Hackman L.E., 1983,"Use of Admixtures in Refractory Concretes", Bull. Amer. Ceram. Soc.62 (9), pp. 1019-25.

[6] Rodger, S.A. and Double, D.D., 1984,"The chemistry of Hydration of High Alumina Cement in the presence of Accelerating and Retarding admixtures", Cement and Concrete Research. 14, pp. 73-82.

[7] Hommer, H., and Seyerl, J.V., 2008,"Impact of Dispersant Structure on Workability and Green Strength Development of LCC at Different Temperature", Proceedings of the 51st International Colloquium on Refractories, 15-16 October 2008, Aachen, Germany, pp. 92-94.

[8] Pivinskii, Yu.E., 1990,"Refractory Concretes of a New Generation :Low-Cement Concretes and Castable Vibration-Treated (Vibrocast) Thixotropic Refractory Bodies-A Review", UDC 666.974.2:66.043:1:693.693.546.4, Plenum Publishing Corporation, 1991, All-Union Institute of Refractories (VIO), Translated from Ogneupory, No.7, pp. 1-10.

[9] ASTM Standards D 422/2007, 2008,"Method for Particle-Size Analysis of Soils", Annual book of American Society for Testing of Material (ASTM), U.S.A., Vol. 4.08.

[10] ASTM Standards C 1446/2007, 2008,"Standard Test Method for Measuring Consistency of Self-Flowing Castable Refractories", Annual book of American Society for Testing of Material (ASTM), U.S.A., Vol. 15.01.

[11] ASTM Standards C 20/2007, 2008,"Standard Test Methods for Apparent Porosity, Water Absorption, Apparent Specific Gravity, and Bulk Density of Burned Refractory Brick and Shapes by Boiling Water", Annual book of American Society for Testing of Material (ASTM), U.S.A., Vol. 15.4.

[12] BS EN Standards 993-5/2000,"Methods of test for Dense Shaped Refractory Products-Part5: Determination of Cold Crushing Strength", British Standard.

[13] ES 6444-1/2007,"Determination of Refractories Resistance for Thermal Shock (Spalling), Part (1): Air Quenching Method", Egyptian Organization for Standardization and Quality, Cairo, Egypt.

[14] ASTM Standards C 179/2009, 2010,"Standard Test Methods for Drying and Firing Linear Change of Refractory Plastic and Ramming Mix Specimens", Annual book of American Society for Testing of Material (ASTM), U.S.A., Vol. 15.01.

[15] Chesters J.H., 2005,"Refractories Production and Properties", Third edition, The Iron and Steel Institute, Carlton House Terrace London, Chapman & Hall.

H.B.G. Ghazal

The High Institute of Engineering, Shorouk, Cairo, Egypt
Table 1: Correlation table for % water added
for bauxite based mixes containing citrate.

 %Cement % citrate % Water

%Cement 1
% citrate 0 1
% Water 0.14437 -0.94254 1

Table 2: Correlation table for Cold Crushing Strength
for bauxite based mixes containing sodium citrate.

 % Cement % Citrate Time CCS

% Cement 1
% Citrate 0 1
Time 0 0 1
CCS 0.091149 0.136458 0.877914 1
COPYRIGHT 2011 Research India Publications
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2011 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Ghazal, H.B.G.
Publication:International Journal of Applied Chemistry
Date:Jan 1, 2011
Previous Article:The viscoelastic properties of Trinidad Lake Asphalt-used engine oil blends.
Next Article:Synthesis and characterization of some mixed ligands complexes from o-phthalic acid as primary ligand.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |