Journal of Minerals & Materials Characteri zation & Enginee r ing, Vol. 8, No.9, pp 667-678, 2009
jmmce.org Printed in the USA. All rights reserved
667
Development of Insulating Refractory Ramming Mass from Some Nigerian
Refractory Raw Materials
O. A. Olasupo1 and J. O. Borode2
1 Engineering Infrastructure Department,
National Agency for Science and Engineering Infrastructure,
PMB 391, Garki, Abuja, Nigeria solayode@yahoo.com
2Metallurgical and Materials Engineering Department,
The Federal University of Technology, Akure, Nigeria borode202@yahoo.co.uk
Corresponding Author: O. A. Olasupo solayode@yahoo.com
Abstract
Insulating refractory ramming mass was developed from suitable Nigerian refractory raw
materials. Rammed samples from several ratios of clay, silica, mica, bentonite and calcium
aluminate cement (Durax) were prepared using the American Foundrymen Society (AFS)
standard rammer. They were thereafter tested for such properties as apparent porosity, volumetric
firing shrinkage, cold compression strength, green compression strength, loss on ignition,
thermal shock resistance and refractoriness. Results indicate that eight ramming cycles were just
enough for the production of the ramming masses. Two optimal ratios obtained from the
experiments have a refractoriness of 1500, good compression strength and excellent thermal
shock resistance. They are therefore recommended for lining of rotary furnaces and crucible
furnaces for the melting of ferrous and non-ferrous alloys. It could be concluded that the
ramming mass serve as a viable alternative to foreign ramming mass at the same temperature
application.
1. INTRODUCTION
A refractory material is one that can withstand the action of abrasive or corrosive solids, liquids
or gases at a high service temperature. They are classified according to shapes and chemical
compositions. Classifying according to shape, there are shaped bricks and the monolithic
refractories. Monolithic refractories are refractory materials that are single piece casts. They have
advantages of eliminating joints which is an inherent weakness of refractory bricks linings. They
have faster application methods, requires no special skill for installation and easy to transport
668 O. A. Olasupo and J. O. Borode Vol.8, No.9
and handle. They also have better spalling resistance and have greater volume stability, with
considerable heat savings than the refractory bricks.
Monolithic refractories are put into place using various methods, such as ramming, casting,
gunning, spraying, and sand slinging. Ramming requires proper tools and is mostly used in cold
applications where proper consolidation of the material is important. Ramming is also used for
air setting and heat setting materials. Because calcium aluminate cement is the binder, it will
have to be stored properly to prevent moisture absorption. Its strength starts deteriorating after 6
to 12 months (UNEP, 2006).
Some earlier works on Nigerian refractories include that of Borode et al (2000) on the suitability
of some Nigerian clay as refractory raw materials. Hassan (2001) worked on the effect of silicon
carbide on some refractory properties of Kankara Clay. Nnuka and Apeh, (1991) worked on the
characterization of Ukpor clay deposits while Nnuka and Okunoye, (1991) worked on the
industrial potentials of the same deposit. Onyemaobi (2004), worked on the assessment of some
refractory properties of some local clays for foundry usage. Ugheoke Onche, Namessan, and
Asikpo (2006), assessed the property optimization of kaolin-rice husk insulating fire – bricks.
All these works have been centered on the production of refractory bricks from the raw materials.
This present work however serves to develop a monolithic ramming mass from some Nigerian
refractory raw materials using bentonite and calcium aluminate cement (Durax brand from
Vesuvius Inc., USA) as binders.
2. EXPERIMENTAL PROCEDURE
The refractory materials used for the various mixes are clay from Ikere-Ekiti, Ekiti State,
southwest Nigeria, silica (coarse-grained), mica from Ijero Ekiti, southwest Nigeria, Bentonite
and calcium aluminate cement (Durax brand from Vesuvius Inc. USA) as binders.
2.1 Equipment
The main equipment used for the preparation and evaluation of the ramming mass properties
includes jaw crusher, ball mill, sieve shaker, sand mixer, AFS ramming machine, Vecstar Electric
furnace, Universal sand testing machine and cold compression strength tester.
2.2 Methodology
The as-received clay was dried, crushed in a laboratory pulveriser and ground in a ball mill. It
was later sieved through a 30m sieve and the oversized that is predominantly silica was rejected.
The silica was washed and dried to remove impurities that may lower the refractoriness of the
ramming mass final mixes. The mica was washed and dried to remove organic impurities and
clay. Milling in a ball mill was done. The mica was then sieved through a 300µm sieve.
2.2.1 Chemical analysis of the raw materials
Vol.8, No.9 Development of Insulating Refractory Ramming Mass 669
669
The chemical analysis of the raw materials was done using the atomic Absorption spectrometer
(AAS) and the results obtained are presented in Table 1.
2.2.2 Particle size analyses
The particle size analysis of silica was done using the American Foundrymen Society (AFS)
Standard. The results of the analysis and the particle size distribution of other materials are
presented in Table 2.
2.2.3 Test sample preparation
The various mixes prepared were mixed using the laboratory sand mixer. Each composition is
first mixed in the dry state for a period of 10mins to facilitate thorough mixing. Water was then
added gradually to the dry powder and the mixing continued for another 20mins. At this point, no
dry powder was left. The mixed sample was then withdrawn form the machine. Ramming of the
sample was made to the AFS specification of Φ50 x 50mm height. Samples were dried in a
laboratory oven at a temperature of 120 for 130mins and then left to cool overnight. Firing of
the samples were done at the rate of 5/min to a temperature of 1100.
2.3 Testing of Samples
2.3.1 Shrinkage test
This was done to determine the volumetric firing shrinkage of the fired samples. The procedure
involved taking the dimensions of the sample before and after firing. The difference in the
dimension was calculated from the expression:
  1  2
1 100%
Where  =volumetric shrinkage
1=initial volume
2=final volume
2.3.2 Appar ent por osity
The weight of each fired sample was taken and recorded as D. Each sample was immersed in
water for 6hrs to soak and weighed while been suspended in air. The weight was recorded as W.
Finally, the specimen was weighed when immersed in water. This was recorded as S. the
apparent porosity was then calculated from the expression:

 x 100%
670 O. A. Olasupo and J. O. Borode Vol.8, No.9
2.3.3 Loss on ignition (LOI)
This determines the percentage weight reduction of the total weight of the sample. The
percentage loss in weight of each sample was determined by the difference between the weight
before firing and the weight thereafter. This represents the loss on ignition at that temperature.
The loss on ignition was calculated from the equation:
 1  2
1 100 %
Where  =Loss on ignition
1=weight of sample before firing
2=weight of sample after firing
2.3.4 Cold compression stre ngth
This was done to determine the compression strength to failure of each sample, an indication of
its probable performance under load. Each sample was placed between two plates of the
compression strength tester. This was followed by the application of a uniform load to it. The
load at which a crack appears on the sample was noted and the cold compression strength  is
calculated from the equation:
   
    2
2.3.5 Thermal shock resistance
Each sample dimensioning Φ50mm ±0.5mm and 50mm±0.5mmheight was placed in an
electrically heated furnace to attain the test temperature of 1100. Each sample was then
withdrawn from the furnace and held for 10minutes. The procedure was repeated until an
appearance of a crack was visible. The number of cycles necessary to cause a crack was recorded
for each of the samples and taken as a measure of its thermal shock resistance.
2.3.6 Refractoriness
This was done to determine the temperature at which each test sample would fuse. Each test
sample was placed in the furnace and the temperature was raised to 1000. The sample was then
observed to check for fusion. The process was repeated by increasing the temperature at 50
interval until fusion was observed.
In the course of the experiment, the following parameters were studied:
(a) The effect of addition of silica on the drying properties of Ikere-Ekiti clay
(b) The effect of addition of mica on the green compression strength and drying properties of the
Vol.8, No.9 Development of Insulating Refractory Ramming Mass 671
671
silica-clay mixture
(c) the effect of binders (bentonite and Durax) on the properties of an optimum clay-silica-mica
mix
3. RESULTS
(a) The result of the chemical analysis of the raw materials is presented in Table1.
Table 1: Chemical Analysis of the Raw Materials
Chemical
Compound Composition
(%)
Ikere-Ekiti
clay Silic
a
Ijero-Ekit
i
mica DuraxBentonite
SiO2 46.28 95.2
6
50.87 15.4355.4
Al2O3 27.71 1.2233.50 34.1323.1
Fe2O3 5.84 0.640.184 5.64 4.4
CaO 1.81 0.011.57 21.820.2
K2O 0.85 0.00
4
0.85 0.85 3.0
MgO 3.65 1.221.30 2.18 2.5
Na2O 0.5 0.690.95 0.13 3.0
MnO 0.74 0.350.57 1.87 0.02
LoI 12.6 0.5610.20 17.677.8
Others 0.02 0.04
6
0.006 0.28 0.58
Total 100 100100 100 100
(b) The results of the sieve analysis and the particle size distribution of the raw materials is
presented in Table 2.
Table 2: Particle Sizes of Raw Materials
Raw Material Particle Size
Ikere Ekiti Clay <300
μ
m
Ijero mica <300
μ
m
Durax <300
μ
m
Bentonite <300
μ
m
Silica AFS GFN 3.81
672 O. A. Olasupo and J. O. Borode Vol.8, No.9
(c) The results of the various tests carried out to determine the characteristics of the test
specimens are presented in Figs. 1 to 9.
(d) The result of the thermal shock resistance and the refractoriness values of samples are
presented in Tables 3 and 4.
Table 3: Thermal Shock Resistance of Optimum Compositions at 1100oc
Sample Number of Cycles withstoodComment*
Clay-Silica-mica-bentonite >30 Excellent
Clay-Silica-Mica-Durax >30 Excellent
* Gilchrist, 1977
Table 4: Refractoriness Values of Optimum Compositions
Sample Refractoriness (oC)
Clay-Silica-mica-bentonite 1500
Clay-Silica-Mica-Durax 1500
4. DISCUSSIONS
4.1 The Effect of Addition of Silica on the Properties of Ikere-Ekiti Clay
It was observed that the Ikere-Ekiti clay mixed with 10% water cracked during the drying
process and firing operations at a temperature of 120oC and 1000oC respectively. This is due to
the volumetric shrinkage, which followed the drying and the firing operations. Therefore, to
reduce the shrinkage and avoid cracks, free silica of AFS GFN 3.81 was added in steps of 2%
w/w of silica, keeping the water content constant at 10% (Fig. 1).
The choice of this grain size is due to the fact that it has been established that free silica of finer
grain size lowers the refractoriness of clay by reacting with the alkali and the alkaline earth
oxides in clay due to larger surface area, which will increase the tendency for softening (Adeoye,
0
2
4
6
8
10
12
14
01234567891011
Volumetric Dr y i ng
Shrinkage (% )
Silica (%)
Fig 1: The Effect of Silica Addition on the Volu me tric
Drying Shrinkage of Ikere-Ekiti Clay
Vol.8, No.9 Development of Insulating Refractory Ramming Mass 673
673
1986). It was observed that compositions with 10% silica and above showed no crack. This is
due to the low volumetric shrinkage after firing. The shrinkage was reduced from 11.5 % in 100%
Ikere–Ekiti clay to 5.9% in 90% clay-10% silica mix. Since the 5.9% value is acceptable for
ramming purposes (Chesti, 1986), the 10% silica:90% clay mix was chosen for further
investigations. The choice was arrived at based on its low volumetric change after firing and
because of the fact that higher silica content will lower the cold compression strength of the mix
and present a rough surface to the rammed samples. It has been established that rough surfaces
usually lower the abrasion resistance of monolithic working surfaces (Semler, 2002).
4.2 Effect of Addition of Mica on the Green Compression Strength and the Drying
Behavior of Clay-Silica System
The effect of mica addition of the green compression strength of clay- silica system is given in
Fig 2. It was observed that the mica addition to the clay-silica-system causes progressive
decrease in the green compression strength of the mix.
Also, fine cracks appeared on the surface of samples with mica content exceeding 20%. This
observed behavior is as a result of the inert surface of mica to water film, which tends to cause
cleavage between the mica particles, silica and clay.
Fig. 3 shows the variation of ramming cycle with apparent porosity.
Figs 4 and 5 show the drying properties of the clay-silica-mica system, while Fig. 6 presents the
loss on ignition at different temperatures. The clay-silica-mica (90:10:20) system was therefore
chosen for further investigations.
0
20
40
60
80
100
120
140
8 101214161820222426
Green Compression Strength
(KNm-2)
Mica (%)
Fig.2: The Effect of Addition of M ica on the Green
Compression Strength of 90:10 Clay:Silica System
674 O. A. Olasupo and J. O. Borode Vol.8, No.9
0
0.5
1
1.5
2
2.5
3
0 20406080100120140
Water Loss (g)
Time (s)
Fig. 4: Drying Rate of Clay-Silica-Mica System
0
20
40
60
80
100
120
0 20406080100120140
Cumulative Water Eveporated (%)
Time (s)
Fig. 5: Cumulative Water Loss in Clay-Silica-Mica System
During Dryi ng
0
10
20
30
40
50
01234567891011
Apparent Porosity (%)
Ramming Cycle
Fig. 3: The Effect of Ramming Pressure on the Ap pa ren t
Porosity of Clay Silica-Mica system
Vol.8, No.9 Development of Insulating Refractory Ramming Mass 675
675
4.3 Effect of Addition of Binders on Some Properties of Clay–Silica–Mica System
Measurements were taken to determine the effect of bentonite and Durax on some properties of
clay-silica-mica system (90-10-20). The samples were subjected to eight ramming cycles to
achieve apparent porosity of 23.1%, which is comparable to that of aluminosilicate bricks
(Gilchrist, 1977). Properties such as volumetric firing shrinkage, loss on ignition and apparent
porosity were also determined. The results were presented in Figs. 7, 8, and 9.
It can be seen from Fig. 7 that an increase in bentonite addition content caused a slight decrease
in volumetric firing shrinkage. This is due to the small particle size of bentonite, which causes
closely packed aggregate resulting in reduction in firing shrinkage as bentonite content increases.
On the other hand, increase in the amount of Durax initially led to decrease in volumetric firing
shrinkage and thereafter remains constant.
0
2
4
6
8
10
12
400 500 600 700 800 9001000110012001300
Loss on Ignition (%)
Temperature (oC)
Fig. 6: Loss on Ignition of Clay-Silica-Mica System with
Temperature
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
0 2 4 6 8101214
VolumetricFiringShrinkage(%)
Binder(%)
Fig.7:EffectofAdditionofBindersontheVolumetric
FiringShrinkageofClaySilicaMicaSystem
Bentonite
Durax
676 O. A. Olasupo and J. O. Borode Vol.8, No.9
Fig. 8 presents the effect of binders on the loss on ignition of clay-silica-mica system. As can be
seen from the curves, the loss on ignition remains fairly constant with increase in binder content.
For the same percentage addition of Durax and bentonite, bentonite addition gave a higher value
of loss on ignition.
From Fig. 9, an increase in the amount of additives caused a corresponding increase in apparent
porosity value of the samples. This is due to the small particle size of the additives, which caused
a close packed aggregate in the structure of the rammed samples.
5. CONCLUSION
From the tests conducted, two optimal ratios of ramming mass could be recommended for use in
lining melting furnaces for cast iron production such as rotary and crucible furnaces. These
materials can also find applications in ceramic kilns and re-heating furnaces operating below
1500oC. They could serve as a substitute for ramming mass in their category.
0
5
10
15
20
25
0 2 4 6 8101214
ApparentPorosity(%)
Binder(%)
Fig.9:EffectofAdditionofBindersontheApparent
PorosityofClaySilicaMicaSystem
Bentonite
Durax
0
5
10
15
20
02468101214
LossonIgnition(%)
Binder(%)
Fig.8:EffectofAdditionofBindersontheLosson
IgnitionofClaySilicaMicaSystem
Bentonite
Durax
Vol.8, No.9 Development of Insulating Refractory Ramming Mass 677
677
References
Adeoye, M. O., (1986): Effect of Microstructures on the Thermal Properties of Refractory
Materials, Unpublished M. Sc. Thesis, Obafemi Awolowo University, Ile-Ife.
Borode, J. O., Onyemaobi, O. O., and Omotoyinbo, J. A. (2000): Suitability of some Nigerian
Clays as Refractory Raw Materials, Nigerian Journal of Engineering Management, Vol.3, pp14-
18.
Chesti, R. A. (1986): Refractories Manufacture, Properties and Applications, Eastern Economy
Edition, Prentice-Hall, pp8-110.
Energy Efficiency Guide for Industry in Asia ©UNEP (2006), Thermal Energy Equipment:
Furnaces and Refractories www.energyefficiencyasia.org accessed on-line 23rd June, 2008.
Gilchrist, J. D., (1977): Fuels Furnaces And Refractories, Pergamon press Ltd. London.
Hassan, S. B. (2001): Effect of Silicon Carbide on Some Refractory Properties of Kankara Clay,
proceeding on Nigerian Metallurgical society.
Onyemaobi, O.O. (2004): Assessment of Some Refractory Properties of Some Local Clays for
Foundry Usage, Inter. Res. Jour. in Engr. Sc. & Tech (IREJEST), Vol.1, No.2, 2004, 105-114.
Semler, C. E., (2002): Refractories Review; The Importance of Microstructure,
www.ceramicindustry.com
Ugheoke, B. I., Onche, E. O., Namessan, O. N. and Asikpo, G. A., (2006): Property Optimization
of Kaolin - Rice Husk Insulating Fire – Bricks, Leonardo Electronic Journal of Practices and
Technologies, Issue 9, July-December 2006,p. 167-178 http://lejpt.academicdirect.org accessed
on-line 23rd June, 2008.