Preparation and characterisation ofself-flowing refractory material containing971U type microsilica C. Go ¨g ˘ tas ¸ 1 , N. U ¨ nlu ¨* 1 , A. Odabas ¸ ı 1 , L. Sezer1 , F. C ¸ ı nar 2 , S ¸ . Gu ¨ner3 , G. Go ¨ller1 and N. Eruslu 1 A newer composition of self-flowing low cement brown fused alumina castable containing 971U type microsilica was developed . Opt imum flow characteristics wer e achieved usi ng wat er addition of 4?6 wt-%. This batch was sintered at different firing temperatures up to 1500 uC. To unders tand the effe ct of bot h firi ng temperatures and correspon ding pha ses, the present cas table was charac ter ised in ter ms of X-ray diffrac tion , sca nning electr on microscopy, bulk density (BD), apparent porosity (AP), water absorption (WA), cold crushing strength (CCS) and self-leveling flowability. The results revealed that gradual increase in firing temperature from 1100 through 1500uC caused low AP and WA, and high CCS properties due to the densification of the castable. Keywords:Refractory material, Alumina castable, Firing temperature, Microsilica, Characterisation, Self-flowing Introduction Improvements in refrac tory materials have signifi cant ef fect on the most of sc ie nti fic an d techno lo gical developments of the past 100 ye ars, especially on iron an d st eel indu st ry , pe tr oc he mi cal an d othe r industries. 1–3 Over the last three decades, achieving uniform microstructure with low porosity, high strength and lower cement content has led the classification of the ref racto ry cas tab les , as the follo win g, convent iona l cast abl e (CaO .2?5%), lo w ceme nt cast ab le (2 ?5%. CaO.1?0%) , ult ral ow cement cas tab le (1 ?0%.CaO. 0?2%) and cement free castable (CaO,0?2%). 1,4,5 Beside thes e deve lopments , self -flo win g refr acto ry cast abl es (SFCs) with a consistency after mixing that allows them to flow and degas without application of vibration, have joined into the family of refracto ry castable s. 5–8 Successful self-flowability and castable packing density in refrac tory castables have been attribute d to existe ncy of the fine particles (fillers) in appropriate amounts. 9–11 Alt hou gh the ben efit s (i.e . high mech ani cal str ength, cre ep res ist anc e and the rma l sho ck pro per ties ) of the well known refractory castable filler, microsilica (silica fume) have been evaluated by several researchers, 12 the development of new type micr osi lica fill er, the ref ore, producing of new type self-flowing refractory castables with improved flow characteristics are underway. This study focuses on develo ping self-flowin g brown fused alumina castable containing 971U-type microsilica. Experimental The starting raw materials used in this study and the compo sition of the studied refra ctory castable are listed in Tables 1 and 2, respectivel y. Experiment s have shown that 4?6 wt- % water is sati sf actory to obtain a self - flowing well dispersed castable system. The preparation of the castable suspension has been mentioned detailed in our previous work. 13 Flow value measurements were carried out by pouring the castable suspension into the tru ncated flow cone as des cri bed in ASTM standa rd C230. For the physical tests such as apparent porosity (AP), bulk den sity (BD), wat er abs orption (WA) and cold crushing strength (CCS), the samples were cast by a si mple tapping tec hnique wi th cube type moul ds (50 mm), then cured for 24 h at room temperat ur e (25uC). After demoulding, these samples were dried at 110uC for 24 h then fir ed at di ff erent temperat ures (1000, 1200 and 1500uC) for 2 h and furnace cooled. The apparent porosity (AP), bulk density (BD), and water abso rp ti on (WA) of th e sa mple s were de te rmin ed acc ord ing to DIN 51056. The cold crushing strength (CCS) was measured by a Naber Model press according to DIN 51067. Seven samples were tested for each different temperature and their standard deviations were cal cula ted . Aft er the cold crushing tes ts (CCS), the eval uati on of the ph ase fo rmat ion of th e cast ab le samples was det ermine d by X-r ay dif fracti on (XRD) ana lys is. The fra ctured sample s (Fi g. 1) wer e crushe d a nd s ie v ed unde r 90 mm for XRD analysis. A 1 Materials Science and Metallurgical Engineering Department, Faculty of Chemi stry -Met allur gy, Istanbul Technical Univ ersi ty, 34469 Masl ak, Istanbul, Turkey 2 Istanbul Technical University, Prof. Dr. Adnan Tekin High Technological Ceramics and Composites Research Center , 34469, Maslak, Istanbul, Turkey 3 Refsan Refrak ter ve Insaat Ltd. Sirketi , Sultan Selim Cad. Ucar Sok. No. 10/A, 34330, 4. Levent, Istanbul, Turkey *Correspondence author: [email protected]2010 Institute of Materials, Minerals and MiningPublished by Maney on behalf of the Institute Received 4 December 2008; accepted 8 March 2009 6 Advances in Applied Ceramics 2010 VOL 1 09 NO1 DOI 10.1179/174367609X4 22199
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Figure 2 shows the XRD results of the studied castablesat different firing temperatures between 110 and 1500uC.
As shown in Fig. 2a, the castable dried at 110uC for 24 h
consists of corundum (Al2O3, a rhombohedral structure
with lattice parameters a5b50?4758 nm, c51?2991 nm,
PDF no. 100173).14 This result indicates the excess of
alumina in the present castable composition. In additionto the phases of the corundum observed at 110uC
(Fig. 2a), one silicon oxide phase (SiO2, a cubic structure
with lattice parameters a5b5c50?713 nm, PDFno. 270605),15 and the phases of albite [Na(Si3Al)O8, a
triclinic structure with lattice parameters a50?8165 nm,
b51?2872 nm, c50?7111 nm, PDF no. 100393],16
anorthite (CaAl2Si2O8, a triclinic structure with
lattice parameters a50?81756 nm, b51?2872 nm, c51?41827 nm, PDF no. 411486)17 were detected from the
XRD pattern of the castable sample fired at 1000uC for
2 h, as seen in Fig. 2b. It should be mentioned here that
since both anorthite and albite phases have same crystal
structure with the lattice parameters very close to each
other, the respective X-ray peaks for both phases are
coincident with each other. Observing one diffraction
peak of the silicon oxide indicates a minor content of
cristobalite. This belongs to the devitrification process of
microsilica particles starting about 1075uC, as reported
by Chakravorty and Ghosh.18 When the firing tempera-
ture increases from 1000 to 1200uC, the peak intensity of
silicon oxide decreases, and there is no trace indicating
the formation of mullite phase, as shown in Fig. 2c.
However, the diffraction peak of silicon oxide disap-
peared at 1500uC (Fig. 2d ). In addition to the similar
Table 2 Composition and self-flowability of studiedcastable
Components Amount, wt-%
Brown fused alumina3–5 mm 161–3 mm 270–1 mm 26Calcined alumina HTM 30 10Rotary bauxite 0–0.063 mm 9Microsilica 971U 7Cement Secar71 5Dispersant D7S 0.05Water 4.6Self-flowability 160
Table 1 Composition of starting raw materials used in this study, wt-%
the number of coarser grades of alumina used, wideningthe particle size range of the formulation.7,11,20,25–29
The bulk density (BD) and apparent porosity (AP) of the studied castable containing 971U type microsilicawere in the range of 3?19–3?24 g cm23 and 7?10–11?25%respectively. Figure 4 shows the behaviour of both BDand AP of the castable as a function of the temperature.The BD of the castable dried at 110uC for 24 h, 3?24¡0?03 g cm23, decreases slightly to 3?21¡0?03 g cm23
after firing up to 1200uC, but increases 3?23¡0?03 g cm23 at 1500uC. Evaporation of water from thecastable creates porosity, which results in decreased BD
and increased AP (11?25¡0?52%) at 1200uC. The higherAP observed at 1200uC for the studied castable indicates
that pores obtained during the water evaporation couldnot be eliminated, even after firing at higher tempera-tures (.1000uC). The AP at 110uC, (10?07¡0?27%), is
similar to that of Chakraborty30 who investigated pro-perly designed low cement castables. However, above1200uC, due to converting most of the metastable liquidof the castable into the mullite, and the densification of
the composition, the AP significantly decreases to 7?1¡0?10% at 1500uC. This kind of decrease of the AP wasreported by Gerotto et al.9 Furthermore, the present AP
a 110uC; b 1000uC; c 1200uC; d 1500uC; e and f higher magnification micrographs at 110uC and 1500uC, respectively; g
EDS pattern at a specific point A in Fig.3 a; h the average composition of the whole matrix in Fig.3 a
3 Images (SEM) of the fractural surfaces of castable samples and EDS patterns
Go gtas et al. Preparation and characterisation of self-flowing refractory material
Adv an ces in Ap pli ed Ce ra mic s 2010 VOL 10 9 NO 1 9
value is smaller than that of the authors’ previous
work13
on similar type brown fused alumina castablecontaining 920D type microsilica. Therefore, it is clear
that several factors such as total water content, initialpacking density (or porosity), consequently, the grainsize distribution, and the both amount and quality of the
microsilica have significant effects on densificationcharacteristics of the self-flowing low cement castables.
Figure 5 shows the relationship between the densifica-tion characteristic of the castable and firing temperature,also verifies the behaviour of both the AP and BD given
in Fig. 4. It appears that the WA increases graduallyfrom 3?10¡0?10% at 110uC to 3?51¡0?20% at 1200uC,but decreases to 2?23¡0?06% at 1500uC. Furthermore,the WA value (2?23¡0?06%) of the present castable
sample at 1500uC is similar to the WA value (y2?0%) of the self-flow ultralow cement high alumina castablecontaining 5 wt-% of microsilica.31
Figure 6 presents the cold crushing strength (CCS) of
the castable samples as a function of the firing tem-perature. The increase in CCS is observed from 924¡18 kg cm22 for 110uC dried samples to 1784¡
160 kg cm22 for the samples fired at 1500uC. Thecastable sample, dried at 110uC for 24 h, is about 0?5
times stronger than the similar type brown fused
alumina castable containing 920D type microsilica, itis also close to that of the self-leveling cement tabularalumina castable with 4?7% water addition.32 Further-
more, the present CCS values are significantly higherthan those of a lightweight low cement self-flowingcastable,33 i.e. 924¡18 kg cm22 for 110uC and 1072¡46 kg cm22 vs 296 kg cm22 at 110uC and 887 kg cm22
at 1000uC. The mechanical characteristics of ceramic
systems containing high alumina contents are related tothe phase formation (i.e. amount of corundum, a glassyphase composition, etc.) and microstructure (crystalsize, and shape, a glassy phase and secondary crystalline
phase distribution, porosity, etc.).34,35 As seen in Fig. 2band c, the anorthite, albite, and minor silicon oxidephases crystallise and gradual increase observed in CCS
indicates the ceramic bond formation. After firing at1500uC for 2 h, the castable exhibits the highest value of
the CCS due to not only the enhanced densification, butalso the formation of elongated needle shaped mullitecrystals which strengthens the structure at high heattreatment temperatures.36
Conclusions
In the present investigation, the authors have prepared aself-flowing low cement brown fused alumina basedrefractory castable containing 971U type microsilica.Water addition (4?6%) has been used to achieve full self-flowing characteristics. The castable sample self-flowedunder its own weight and easily filled intricate shaped
moulds without any vibration. At 110u
C, the matrixconsists of corundum phase. With increasing firingtemperature through 1200uC, the phases of corundum,anorthite, albite and silicon oxide were detected. At1500uC, the mullite formation was observed due to thereaction of silicon oxide with corundum. The investi-gated physicomechanical characteristics exhibited thatseveral factors such as total water content, initialpacking density (or porosity), therefore, the grain sizedistribution, and both amount and quality of themicrosilica have significant effects on densificationproperty of the self-flowing low cement castables.
Acknowledgements
This work was financially supported by the IstanbulTechnical University, Turkey, Research Fund (Project
4 Densification behaviour of castable samples as a func-
tion of firing temperature
5 Water absorption of castable samples as a function of
firing temperature
6 Cold crushing strength of castable samples as a func-
tion of firing temperature
Go gtas et al. Preparation and characterisation of self-flowing refractory material
10 Advances in Applied Ceramics 2010 VOL 1 09 NO 1
No. 30329). N. U nlu would like to thank ProfessorMohan J. Edirisinghe for his helpful discussion. Theauthors wish to thank Mr Huseyin Sezer for his help andcontributions during the SEM analysis of this study.
References1. W. E. Lee and R. E. Moore: J. Am. Ceram. Soc., 1998, 81, 1385–
410.
2. G. Lewis: in ‘Applications for traditional ceramics, refractories’,(ed. S. J. Schneider), 895–918; 1991, Metals Park, OH, ASM
International.
3. A. R. Studart and V. C. Pandolfelli: Am. Ceram. Soc. Bull., 2002,
81, 36–44.
4. L. P. Krietz, R. E. Fisher and J. G. Beetz: Am. Ceram. Soc. Bull.,
1990, 69, 1690–1693.
5. W. E. Lee, W. Vieira, S. Zhang, K. Ghanbari Ahari, H. Sarpoolaky
and C. Parr: Int. Mater. Rev., 2001, 46, 145–167.
6. F. G. Brachet, R. Avis, and B. Clavaud: Proc. Unified Int.
Technical Conf. on ‘Refractories’, The German Refractories
Association (GRA), Aachen, Germany, September 1991, 186–
188.
7. S. K. Das, R. Sarkar, P. K. Mandal and S. Mukherjee: Am. Ceram.
Soc. Bull., 2003, 82, 55–59.
8. P. C. Evangelista and C. Parr: Refrac. Appl. News, 2002, 7, 14–18.
9. M. V. Gerotto, A. R. Studart, M. D. M. Innocentini, V. C.Pandolfelli and S. S. Cabo: Am. Ceram. Soc. Bull., 2002, 81, 40–47.
10. B. Myhree and A. M. Hundere: Proc. 39th Int. Colloquium on
‘Refractories’, Verlag Stahleisen, Aachen, Germany, September
1996, 184–188.
11. A. R. Studart, W. Zhong and V. C. Pandolfelli: Am. Ceram. Soc.
Bull., 1999, 78, 65–72.
12. M. V. Gerotto, A. R. Studart, R. G. Pileggi and V. C. Pandolfelli:
Am. Ceram. Soc. Bull., 2000, 79, 75–83.
13. C. Gogtas, A. Odabası, L. Sezer, F.Cınar, S. Guner, N. Eruslu and
N.U nlu: J. Ceram. Proc. Res., 2007, 8, 324–330.
14. Powder Diffraction File, Card No. 10-0173: Joint Committee on