My research activities are mainly in refractories, their microstructures, postmortem analysis and characterization, synthesis of ceramic powders by mechanochemical techniques. Especially, MgO-C, MgO-chromite, spinel bricks and basic self flow castables are the subjects for research. Recently, we started working also on low temperature synthesis of spinel and cordierite via mechanical grinding. This technique involves the storage of strain energy in the powder to facilitate the initiation of phase transformation at a lower temperature. Another project that we have recently started deals with the development of dense alumina ceramic tiles from mixtures of domestic aluminas to achieve high density tiles. Apart from these, use of AI (Artificial Intelligence) techniques like ANNs (Artificial Neural Networks) for modeling of ceramic plant data is under focus. We have produced two refereed journal articles from ANN modeling in 2003. Work is also underway on the historic mortars and plasters. We are working on the pozzolanic activities of heated clay minerals and their use in conservation of historic building.
Ongoing Research
Microstructural Characterization of Industrial Chromite and Spinel Cement Kiln Refractories with Emphasis on the Iron-rich Rims
Yener Mercanköşk and Sedat Akkurt
Abstract
Magnesia-chromite (MgO + MgO.Cr2O3) and magnesia-spinel (MgO + MgO.Al2O3) refractory bricks are widely used in the high temperature zones of rotary cement kilns. Although the former is now being replaced by the latter and by the dolomitic grades due to the carcinogenic Cr+6 content, it is still used in most cement kilns in Turkey. Their microstructures are important because the size, shape and distribution of periclase grains, chromites and the quality of their bonding phases significantly affect their service performance. The purpose of this study was to characterize the microstructures of industrial brick samples to develop a protocol to compare different products e.g. for evaluation as replacement bricks. In some of the chromite containing bricks iron-rich rims were observed, while a domestic brick with similar chemistry had no such feature. These iron-rich rims were examined using SEM-EDS. It was found that the counter-diffusion of Fe+2 and Mg+2 were responsible for their formation. Exsolved chrome-spinel was widely observed in the microstructures of chromite bricks. Magnesia-spinel bricks were found to contain low melting calcium aluminates as bond phases in the microstructure, posing a threat to service performance. Portmortem microanalysis of industrially used bricks revealed alkali attack in addition to creep as main destruction mechanisms for brick. Traces of elements like zinc were observed and thought to originate from the use of waste derived fuels.
Low Temperature Synthesis of a-alumina, Cordierite and Spinel Powders by Mechanical Grinding
Emre Yalamaç and Sedat Akkurt
Abstract
Low temperature synthesis of a-alumina, cordierite and spinel powders has been attracting attention in recent years due to new potential applications. The use of mechanochemical methods to achieve partial or complete structural disorder as a tool to lower the sintering temperatures has also been increasingly reported. Traditionally, cordierite is synthesized by mixing proper amounts of various combinations of alumina, clay, talc (or sepiolite as a source of MgO) and magnesium hydroxide (Mg(OH)2). In this study, however, their mixture combinations are programmed using statistical experimental design methods via e.g. the D-optimal mixture design. The purpose was to produce the ideal composition in a way to accomodate variations in the raw material compositions. Intense mechanical grinding allowed the lowering of synthesis temperatures. Detailed microstructural characterization of the synthesized pellets was performed using SEM-EDS. XRD, TGA, DTA and FTIR were used for analysis of powder products. Each stage of synthesis was analyzed using these techniques. Temperatures as low as 950oC were able to produce a dense structure. Similar results were obtained for alumina and spinel synthesis. The former has been shown to convert into alpha phase as low as 900oC. The latter was produced by intense milling of a mixture of Mg(OH)2 and Al(OH)3.
Production and Characterization of Bauxite-containing Electrical Porcelains
Mücahit Sütçü and Sedat Akkurt
Abstract
High voltage electrical porcelain bodies have recently been produced by the use of bauxite as a filler to replace the more expensive alumina additives. In the past, quartz was used as a filler but was later replaced by alumina due to improved mechanical properties. The use of bauxite not only offers the potential for reduced cost, it also provides comparable mechanical properties to alumina. These porcelains are required to have high mechanical strength, low porosity, dense structures and also to perform reliably even under the harshest weather conditions. Therefore, microcracks must not occur because of differences of thermal expansion coefficients of the phases (corundum/glassy phase or quartz/glassy phase) in the structure. Free quartz particles must be minimal in the matrix.
In this study, various combinations of bauxite as filler, clay as binder and feldspar as flux were used. Using statistical experimental design program the mixtures were prepared. Pellets were pressed from these mixtures and soaked at different temperatures (1300-1500° C) and linear shrinkage, porosity and density were measured. Optical microscope and SEM-EDS were used for microstructural analysis. The phase development was closely monitored via SEM-EDS. Compositional changes on microscale were tracked to understand the phase development. In the porcelain bodies, the formation of mullite, corundum and glassy phase were observed. X-Ray Diffraction was used to confirm phase formations as mullite, corundum, etc. Successful compositions were developed to match the initial specifications. Effect of intense grinding is also tested and the preliminary results are promising in that the sintering temperature can be significantly reduced.
Investigation of the Pozzolanic Properties of Bricks Used in Horasan Mortars and Plasters in Historic Buildings
Hasan Böke, Sedat Akkurt, Başak İpekoğlu
Abstract
Brick – lime mortars and plasters known as Horasan in Turkey, Surkhi in India, Homra in Arabic countries and Cocciopesto in Romans have been used since ancient times due to their hydraulic properties. These materials are composed of a mixture of finely ground brick and lime which form a hydraulic compound if the crushed bricks are pozzolanic.
Historical mortars and plasters need conservation due to their deterioration problems. During their conservation, the new mortars and plasters must be compatible with the existing ones. Modern bricks used in the manufacturing of Horasan mortars and plasters are seldom pozzolanic because they are fired at high temperature and they have low clay contents. Therefore, the correct choice of bricks used in the preparation of new Horasan mortars and plasters is important to manufacture new intervention mortars.
In this work, to define the properties of bricks used in the preparation of new intervention mortars or plasters, bricks used in sound historic Horasan mortars and plasters collected from historic bath buildings have been examined. Brick powders and fragments were separated from calcium carbonate with dilute hydrochloric acid. The mineralogical and chemical composition, microstructures of the separated brick fragments were determined by using X-ray diffraction (XRD), optical microscopy and scanning electron microscopy (SEM) coupled with EDS (EDAX). Pozzolanicity of the brick powders were determined by mixing them in a saturated calcium hydroxide solution and by measuring the differences in electrical conductivity.
Ettringite Formation in Historic Bath Brick-Lime Plasters
Hasan Böke and Sedat Akkurt
Cement and Concrete Research, In press, November 2003
Abstract
In this study, hydraulic brick-lime plaster samples from selected Ottoman baths were examined to characterize their technology and the appropriateness of their use in baths. The samples that were historical repair-plasters were found to have deteriorated despite being exposed to the same environment with the original more sound samples. This was investigated by comparing their raw materials composition and the pozzolanic activity differences of the brick powders used in the plasters. Although the results showed no significant differences, considerable amount of ettringite crystals was detected in the historic repair ones by XRD, FTIR and SEM-EDS analysis. Due to the ettringite formation, the repair plaster may have lost its integrity owing to the expansion generated by the growth of ettringite crystals in the plaster. In this study, ettringite formation has been discussed in relation to hydration reaction products of lime – brick mixture, possible sources of gypsum and the climatic conditions of historic bath building.
Keywords: Historical Brick-Lime Plaster, Ettringite, Gypsum, Deterioration
Prediction of the Slag Corrosion of MgO-C Ladle Refractories by the Use of Artificial Neural Networks
Sedat Akkurt
Abstract
Artificial neural networks (ANN) are used for prediction of the amount of corrosion of MgO-C ladle refractories in molten steel-refining slags. Laboratory slag corrosion test data generated in a previous project are fed to a multilayer feed forward backpropagation (MFFB) learning algorithm for training of the model. Genetic algorithms (GA) are employed for selection of training data and testing data. Based on the training data a GA-ANN model of the amount of corrosion is created. Testing of the model is also performed and successfully low average error levels (7.6%) are achieved. This model is then used to predict the response of the slag-corrosion system to different values of the factors affecting the corrosion of bricks at high temperatures. Factors used for modelling were exposure duration, exposure temperature of slag-brick contact and CaO/SiO2 ratio of the slag. Model results provided the potential for selection of the best experimental conditions for avoiding the factor combinations that may accelerate corrosion.
Keywords: Artificial Neural Network; MgO-C Refractory; Corrosion
Characterization of Composite MgO-C Refractory Bricks as an Initial Step to Understanding Corrosion
S.Akkurt* and H.D.Leigh **
Journal of the Canadian Ceramic Society, Vol:78, 9-18, Fall 2002
* Mechanical Engineering Department, Izmir Institute of Technology, Urla 35437 Izmir, Turkey
** Ceramic and Materials Engineering Department, Clemson University, Clemson, SC, 29634-0907, USA
Abstract
A research project aimed at understanding the corrosion of MgO-C ladle refractories by secondary steelmaking slags was performed. As part of this project three types of industrial refractory brick specimens were characterized before conducting the corrosion tests at high temperatures. XRD, DTA, TGA, SEM, OM, TSP (Thin section petrography), screen analysis were used to characterize the bricks both chemically and microstructurally. The results indicated that the brick coded D was prone to corrosive attack owing to its poor microstructure and chemistry. In addition, the low dihedral angle of MgO crystallites in MgO grains confirmed these findings.
Keywords: Characterization; Ceramic; Composite; Refractory; MgO-C brick; Graphite
The Use of Experimental Design in the Study of the Corrosion of MgO-C Ladle Refractories
S.Akkurt and H.D.Leigh
American Ceramic Society Bulletin, In Press, Vol:82, May 2003
Abstract
Industrial refractory bricks were tested in the laboratory for their corrosion behavior at high temperatures under a protective atmosphere of argon or argon+CO(g). Statistically designed set of experiments provided a polynomial equation of the model for corrosion.
The Use of GA-ANNs in the Modelling of Compressive Strength of Cement Mortar
Sedat Akkurt, Serhan Ozdemir, Gokmen Tayfur and Burak Akyol
Cement and Concrete Research, In Press, July 2003
Abstract
In this paper, results of a project aimed at modelling the compressive strength of cement mortar under standard curing conditions are reported. Plant data are collected for six months for the chemical and physical properties of the cement that are used in model construction and testing. The training and testing data are separated from the complete original data set by the use of Genetic Algorithms (GA). A GA-ANN model based on the training data of the cement strength was created. Testing of the model was also done within low average error levels (2.36%). The model was subjected to sensitivity analysis to predict the response of the system to different values of the factors affecting the strength. The plots obtained after sensitivity analysis indicated that increasing the amount of C3S, SO3, surface area and LSF (Lime Saturation Factor) led to increased strength within the limits of the model. C2S decreased the strength while C3A decreased or increased the strength depending on the SO3 level. Because of the limited data range used for training, the prediction results were only good within the same range. The utility of the model is in the potential ability to control processing parameters to yield the desired strength levels and in providing information regarding the most favourable experimental conditions to obtain maximum compressive strength.
Keywords: E- Modelling, C- Cement Strength, Artificial Neural Networks; Genetic Algorithms
The Use of GA-ANNs in the Prediction of Germanium Recovery from Zinc Plant Residues
Sedat Akkurt, Serhan Ozdemir
and Gokmen TayfurIMM Transactions, Section C,
Mineral Processing and Extractive Metallurgy, Vol:111, No:3, C129-C134, March 2003
Abstract
A Multilayer Feed Forward Backpropagation (MFFB) learning algorithm was used as an artificial neural network (ANN) tool to predict the extraction of germanium from zinc plant residues. Genetic Algorithms (GA) were utilized for selection of training data and testing data. Based on the training data a GA-ANN model of the germanium leaching system was created. Testing of the model was also performed with good error levels (r-square=0.95). This model was utilized to predict the response of the system to different values of the factors affecting the recovery of germanium by sulfuric acid leaching. Model results provided the potential for the selection of the best experimental conditions for germanium recovery from zinc plant residues by sulfuric acid based leach solutions.
Keywords: Artificial Neural Network; Germanium; Extraction; Zinc; Genetic Algorithms
Investigation of the Effect of Firing Conditions on Corrosion of High Alumina Crucibles by Steelmaking Slags
S.Akkurt and H.D.Leigh
Abstract
A research project aimed at investigating the corrosion of high alumina crucibles made in a set of different conditions is reported. The slag permeability and corrosion characteristics are the response variables in this study, while the type of reactive alumina, firing temperature and soak time are the experimental factors that are investigated. It is found that corrosion test temperature and duration of slag-crucible contact are the strongest factors affecting percent slag penetration.