Search Documents

By P. Schloth, H. van Swygenhoven, N. Chobaut, J-M. Drezet

"In the current trend toward thicker aluminium plates, a major concern is the stress build-up during quenching which causes distortions during machining. Indeed, cooling rates are not high enough, especially at the core of such thick plates, to prevent any precipitation and quench induced precipitates lower the hardening potential. Multi-scale modelling is required when predicting macro-scale stresses after quenching for thick heat treatable aluminium components. The reason is the instantaneous strong coupling between phase precipitation at the nano-scale and material hardening due to precipitation or softening owing to solute depletion at the microscale. For thick parts, quenching intensities decrease when going from the skin to the core of the component, thus introducing a gradient of nanostructure and consequently a gradient of mechanical properties. In addition, large thermally induced deformations lead to high macroscale residual stresses although part of them is relaxed by plastic deformation. These stresses have been measured in water quenched thick plates of 7040 and 7449 aluminium alloys using neutron diffraction and layer removal techniques and the results when compared with a thermomechanical finite element model of quenching highlight the influence of precipitation.IntroductionIn the processing route of heat-treatable aluminium alloys (AA), a critical step is quenching from the solutionizing temperature. High cooling rates are desired, as they avoid formation of coarse precipitates, which reduce mechanical properties. Fast quenching, however, cannot be achieved in the centre of thick components, where the quenching rate can be more than one order of magnitude lower than at the surface. The existence of a thermal gradient between core and surface leads to a differential thermal deformation which is accommodated by elastic and plastic deformation to comply with the compatibility of deformation. Whenever the material yields, especially at high temperature when the yield strength is low, residual stress remains after quenching due to the irreversibility of plastic deformation. After quenching, the plates are in a state of residual stress that causes unacceptable distortions during machining and may reduce service life drastically. Since stress relaxation is negligible during ageing treatments [1 ], agehardened aluminium parts are usually stress relieved prior to ageing. This is typically achieved by applying plastic deformation: stretching from 1.5 to 3 % in the rolling direction (L) can reduce the internal stresses by a factor of approximately 10 [2, 3]. However, complex geometries or thicker components like thick sheets require other approaches. Reducing the quenching rate by using boiling water for instance leads to lower residual stresses [1, 4]."

Jan 1, 2013

BHP Hartley Platinum operates an electric smelting furnace to treat Cu-Ni-PGM concentrate at its plant in Selous, Zimbabwe. The furnace, originally built by others, has experienced problems including severe refractory erosion of the sidewall. Hatch Associates has designed several important improvements to the furnace, including: ? water-cooled copper cooling elements in the slag zone ?air-cooled copper cooling fins at the metal level ?water-cooled copper tapholes ?binding system providing vertical refractory compression ?electrode seals ?new feed pipes Novel construction techniques employed on this project to reduce shutdown duration are discussed.

Jan 1, 1999

By A. A. Hejja

The first part of this study deals with the importance of surface tension of slags produced in the smelting of sulphide concentrates. Theoretically the method of determining the surface tension of slags was laid down by Boni and Derge. The values obtained experimentally for a series of main oxide components were grouped into a useful correlation incorporating experimental constants for the relevant oxides present in the slag, some of these giving the surface tension value for the particular slag. These estimated values were coupled with or incorporated into the slag viscosity plots to show the correlation between the two factors. This is also important from a refractory attack point of view in the furnace operation. In the second part an attempt is made toward the optimisation of slag composition relevant for electric furnace smelting processes. Using the knowledge of the viscosity and electrical conductivity of the slags at the temperature of furnace operation, the mode of optimisation is based on correlations which utilize the trend of the mostly parabolic plots of viscosities and conductivities. The results obtained by this method agree well with the optimum slag compositions being used in established operations derived from years of practical experience for various slags used in the furnaces. However, due to extensive turbulence in suspension and gas injected bath smelting processes, the outlined methods cannot be used for these processes. Also great caution should be exercised in their application for three electrode circular furnaces with separate slag cleaning.

Jan 1, 1997

By Sadegh Safarzadeh, Anne-Marie Suriano, Stanley Howard

The electrodeposition of metal ions from aqueous electrolyte solutions has been a well-known process for the last half century. With numerous industrial applications such as electroplating, electrowinning and electrorefining, the kinetics of electrochemical reactions involved in the process has been extensively researched. Arrhenius-type rate equations, the Butler-Volmer equation, and the empirical Tafel equation are among the most important models to explain the rate of electrochemical reactions. The study of the kinetics of metal deposition is critical to understanding the underlying mechanisms through which a certain morphology of metal deposit is formed on the cathode. The purpose of this paper is to elucidate the influence of experimental parameters such as pH and additives on the rate and the reaction mechanism determining the rate limiting step of the electrodeposition reactions.

Jan 1, 2015

By Carla Lupi, Alessandro Pescetelli

"Spent lead battery recycling still shows few weak points in the achievement of a 100% recycling of different waste matters and by- products arising from the recycling process. Some of said problems are dealing with aqueous streams outgoing the recycling plant, and more particularly, a) the waste waters discarded from the breaking and separation unit, b) purification of the spent electrolyte drained from waste batteries and upgrade to a quality compatible with full re-use by the batteries’ manufacturers, c) purification of the mother liquors discarded in the sodium sulfate crystallization process so to produce a salt of commercial grade. All of the above listed problems have been satisfactorily overcome applying an electrochemical process that, through different process conditions and parameters, is able to remove the harmful impurities that hinder the successful commercialization of such by products, or, at least, that can avoid heavy and expensive treatment plant and operation.IntroductionAlmost 100% of commercial lead and lead alloys produced by secondary lead plants is recycled to the new lead accumulators manufacture. The other by-products and residues produced through the lead recycling process may be classified into two groups:a) matters reusable in the new accumulators production processb) matters produced as a result of various reagents and additives used in the recycling process, to be sold to different market sectorsQuality and saleability of by-products are significant points for the economics of the secondary lead plant operation, because, if said by-products are not saleable, they are just wastes, so offsetting potential revenues, and, at the same time, adding a heavy burden to operating cost, like disposal cost [1]. When looking at a typical, modern lead scrap batteries recovery cycle, it is quite evident that we are still rather far from dealing with a waste-less technology (fig.1).The group a) includes 3 by-products produced through the breaking process: among those matters the polypropylene only usually reaches a standard quality suitable for sale to lead manufacturing sector or to other sectors as well. The other two by-products, electrolyte and separators, do not still reach, through a viable and effective regeneration process, a quality standard sufficient for their re-use in the lead batteries sector ."

Jan 1, 2008

By Shoji Taniguchi, Lifeng Zhang, Kaike Cai

"In the present paper, based on the two-phase (Eulerian-Eulerian) model, the 3D fluid flows in gas-stirred systems, i.e., air-stirred water vessel and argon-stirred liquid steel ladle, are simulated. In the Eulerian-Eulerian two-phase model, gas and liquid are considered to be two different continuous fields. The phases are assumed to share space in proportion to their volume fractions. The exchange between the phases is represented by source terms in conversation equations. Turbulence simulated by the k- e model is assumed to be the property of the liquid phase. The following effects of mathematical treatments and operational factors on the fluid flow are discussed, such as interphase drag force, turbulent model, the size of bubble, and gas injection mode.Some interesting results are derived. Except the interphase drag force, the interphase lift force should be taken into accounted in order to exactly simulate the fluid flow in gas-stirred systems. Suitable turbulent model and drag coefficient have effect on the mathematical simulation. Injecting small bubbles can realize a well mixing flow condition. The distance between the two gas-injection nozzles has effect on the fluid flow."

Jan 1, 2000

By R. H. Rubisov

Sulphuric acid pressure leaching of nickeliferous laterites has attracted considerable attention from the nickel industry during the past five years. The process is especially advantageous for limonitic laterites that mostly contain goethite, because iron precipitates releasing acid and thus rendering low acid consumption. It is also applicable to mixtures of limonites and saprolites, although at higher acid consumption due to magnesium. Effective process design requires the solubility of metals that precipitate during the process to be known. In the present work, determination of metal solubilities is based on a simple speciation program that assumes the presence of only one dominant complex for each metal. The thermodynamic data for the precipitation reactions are extracted from high-temperature experiments with monometallic systems published previously. The validity of the approach is then tested against mixed bimetallic systems, and finally applied to calculate the solubility of aluminium, iron and magnesium in laterite leaching effluents "at temperature". In both cases of limonitic feed and limoniticfsaprolitic blends, the prediction closely follows metal solubilities measured experimentally at temperatures from 230 to 270°C and at terminal he acidities ranging from 10 to 75 g/L.

Jan 1, 1999

By Leo Brewer, Kazuyoshi Shimakage, Mamoru Mitomo, Toshiyuki Nishimura, Yoichiro Uemura, Shinji Hirai

"In the synthesis process of ß -La2S3 via the sulfurization of La2O3 powder using CS2 gas, La2O2S was formed. This occurred at an initial stage of the reaction for sulfurization temperatures below 973K, and ß -La2S3 was also formed at a shorter period for higher sulfurization temperatures. It was found that the oxygen content in ß -La2S3 powders decreased gradually with an increase in sulfurization temperature. On the other hand, the carbon content in these powders increased gradually with an increase in the sulfurization temperature. Though it has been reported that ß -La2S3 was transformed to Y -La2S3 at 1723K± 50K, the sintered compact of double-phase Y -La2S3 and ß. -La2S3 was formed at a rather higher temperature of 1973K when the synthetic powder of ß -La2S3 containing a high amount of oxygen was used. However, phase transformation of -La2S3 to Y -La2S3 progressed preferentially with a decrease in the oxygen content of synthetic ß -La2S3 powder. The Seebeck coefficient of the sintered double-phase Y -La2S3 and -La2S3 was higher than that of the single-phase Y -La2S3, reported previously.Introduction La2S3 exists in three polymorphic forms that are similar to a series of Ce2S3, Pr2S3 and Nd2S3: a, ß , and Y'. The a -La2S3, which is stable at low temperature, is irreversibly transformed to ß -La2S3 of various compositions in the solid-solution range between La2S3 and La3S4, at 923 ± 50K. The -La2S3 may also be an oxysulfide of La10S14O2. Moreover, the ß .-La2S3 phase is transformed to Y -La2S3 ,which has a Th3P4 structure, at 1573 ± 100K. The melting point of Y -La2S3 is 2398 ± 30K where this temperature is similar to those of common refractory materials such as silica and alumina. Therefore, it can be predicted that y -La2S3 which has a high melting point and a thermodynamic stability due to the strong affinity of La for S, would have &high refractory character and chemical inertness in many molten metals. The y -La2S3, colored vivid faint yellow is stable up to 623K in air, where oxidation starts. y -La2S3 is frequently used as pigments for plastics and paints3. Furthermore, Y -La2S3 which is a semiconductor, is considered to be useful as a thermoelectric conversion materials, because it has a high Seebeck coefficient."

Jan 1, 2000

By M. Giralico

Optimized mixer settler designs will now handle flows of over 25,000 GPM. The size and expense of the equipment hamper experimental settler optimization efforts. Computational Fluid Dynamics (CFD) is a timely and predictable tool for settler design development. In the past, CFD has been used to evaluate only the flow field in pumped and auxiliary tanks. A similar numerical analysis is used to evaluate settler flow and phase separation. Phase separation is analyzed using multiple species for the aqueous and organic phase. Experimental validation shows good agreement between the computational and experimental solutions in both flow field and phase separation. The computational model is used to show performance differences between various choices of settler geometry, liquid depth, with variable flow rates. Understanding the differences allows the development of optimized settler designs. System improvements are predicted to lower entrainment levels further and/or allow new smaller footprint designs. This paper focuses on the major design components of the settler and how future technological advances will lead to further improvements in their design. LIGHTNIN?s new designs will handle tomorrow?s production requirements, while maximizing profits.

Jan 1, 2003

By S. P. Song

A computational methodology based on the coupled finite element and boundary methods is developed to predict the free surface shapes of electrically conducting liquids supported by the electromagnetic field. The work is an extension of our previous FEIBE model for induction calculations. The free surface algorithm is developed using the Weighted Residual Method, the very same used for nonlinear finite element formulation. This allows the treatment of field computation and free surface prediction in the same manner, thus making easier the formulation of the magnetically-supported free surface problems. Like other nonlinear algorithms, the boundary/finite element free surface scheme is also iterative owing to the fact that free surface shapes are unknown a priori. But the numerical scheme is robust and efficient and is capable of handling rather complex free surface deformations. Computational algorithms will be given and are tested by comparing with other solutions whenever available. Computed results of magnetically-supported free surface shapes are presented for magnetic levitation under microgravity and floating zone refining processes under normal gravity.

Jan 1, 1996

By D. V. Hughes, J. R. Nyberg

"Since 1985, a comprehensive automation system has been installed at the Outokumpu Oy Zinc Plant, Kokkola, Finland. In leaching and solution purification special sample handling equipment, on-line analyzers and instruments, and a computer-based control system with duplicated supervisory stations have been installed. The system has enabled closer control of the process and more efficient use of labor. Since its installation recovery of zinc has improved, and use of zinc dust, steam, and water have decreased. The system has performed reliably. It includes optimizing programs and process study tools, which facilitate improving process performance. An expert system has been built-up to prompt process personnel to take appropriate action on the basis of process data. In electrolysis a similar system, based on extensive monitoring of the process and of production of zinc cathodes, is used to maximize- current efficiency by maintaining optimum conditions in the process. It is planned that in the future the control systems will be integrated and extended to cover the roasting plant and effluent control.IntroductionThe Zinc PlantThe Zinc Plant of Outokumpu Oy Metallurgical Division at Kokkola, on the West Coast of Finland, has grown from a 90,000 t/a plant founded in 1969 to become, with its present capacity of 170,000 t/a, one of the world's ten largest zinc plants. Especially since the completion in 1988 of a thorough modernization programme, it is also among the world's most modern.The Kokkola plant takes its raw materials from Outokumpu's own mines, Vihanti and Pyhasalmi in Finland, and Tara in Ireland, and from outside sources. Concentrates are roasted in two Lurgi fluid-bed furnaces, and the calcine is leached and iron removed by the Conversion process. Following this copper, cobalt, and cadmium are removed, and zinc is refined by electrowinning. The leaching and purification, and the cell-house are automated under computer-based control, and anode and cathode handling in the cell-house are mechanized and fully automatic."

Jan 1, 1989

By Patrick R. Taylor

This investigation studied the inflight calcination kinetics of nuclear waste simulant using a thermal plasma reactor. A series of experiments were performed to determine the conversion of insoluble aluminum compounds to the soluble compounds and the decomposition of complex organic and nitrite compounds present in the simulant waste. Thermodynamic calculations based on free energy minimization were performed to establish a theoretical basis for this investigation. A kinetic model to describe the conversion of gibsite to sodium aluminate has been proposed. The products of the calcination were characterized using XRD, Gas Chromatography, and chemical analysis by ICP. The results of this investigation indicated 85% gibsite conversion to sodium aluminate, 90% of nitrate decomposition, 52% TOC decomposition and 90% conversion of chromium to soluble compounds. A process is identified for inflight calcination of wastes. Key words: Plasma, waste, simulant, calcination

Jan 1, 1996

By M. Omneya El-Hussaini, M. Galal Abdel Wahab, A. Mohamed Mahdy, M. Fathi El Shahat

"Um-Gheig Pb-Zn representative sample was collected from the main adite of Um-Gheig mine along Red Sea coast in Egypt. Its chemical analysis shows that, it contains 16% Zn, 2.3% Pb, 0.1% Cd and 0.1% B as elements of interest. After H2SO4 agitation leaching, the chemical composition of Um- Gheig Pb-Zn deposit ore residue assayed 2.26% Pb, 1.5% Zn, 0.085 % B and 0.01% Cd. This residue was leached with 20% HCl at S/L ratio 1/3 for 2 hours at 80o C. The chloride leach liquor was found to contain 18 g/l Pb, 13 g/l Zn, 0.1 g/l Cd and 0.75 g/l B with the corresponding leaching efficiencies 79%, 86%, 100% and 88% respectively. Lead was precipitated from the chloride leach liquor using concentrated H2SO4 to obtain PbSO4 with precipitation efficiency 96%. While boron was extracted by the selective resin Amberlite IRA743 with loading efficiency of 80%. On the other hand, Zn and Cd were precipitated using Na2S solution with 86.7% and 100% recovery efficiencies respectively.IntroductionLead-zinc ores occur in Egypt in a few localities distributed along the Red Sea coast between El Qusier and Ras Banas.. The reserve of the ore deposits has been estimated by Soviet and German Experts to be about 900,000 Ton, [1]. Onal et al., [2] recovered Pb from Pb–Zn sample containing 10.17 % Pb and 10.98 % Zn from Aladag in Turkey. El-Sherif, [3] used concentrated HCl to leach lead-zinc ore sample of Um-Sammuki area in Egypt. As had been published by Marchioretto et al. [4], the precipitation of heavy metals with lime, sulfide and caustic soda are the most commonly used economical methods for metal removal. Lead was completely removed from the anaerobically digested sludge by precipitation with NaOH and Na2S at waste water treatment plants. Mahdy et al., [5] had successfully extracted boron from the bittern solution of Qaron Lake by ion exchange resin (Amberlite IRA 743). The extraction of boron was mainly depending upon the pH value. They extracted 80% of boron at pH 5.5, while at pH 3.9 about 74% boron was extracted. Kabay et al. [6] used the column-mode to remove 1g/l of boron from geothermal waste water of the Kizildere-Geothermal Power Station by using Diaion CRB 02 Nglucamine- type chelating resins. In this regard, the removal of lead, cadmium and boron allow the residue to be reclassified as non hazardous material."

Jan 1, 2008

By A. Slotwinski, R. E. Barry, M. M. Menon, P. T. Spampinato

"A coherent FM laser radar based sensor for remote metrology of plasma facing material surfaces in a fusion reactor is described. The sensor is capable of 3-D mapping of surfaces, based on a series of precise range measurements from large distances (up to 20 m). Other features include the ability to operate under adverse environmental conditions involving very high gamma radiation (3xI04 Gy/hr), ultra-high-vacuum (<I0-5 Pa), and high temperature (200°C). The paper describes the basic principles of the sensor, together with experimental results obtained with the sensor. The ability of the sensor for remote rendering of ""visual quality"" images without the need for any external illumination is also illustrated.IntroductionMeasurement of the features of large area surfaces (-1500 m2) to sub-millimeter accuracy, from relatively large distances (20 m), is necessary in the case of the International Thermonuclear Experimental Reactor (ITER -a 1500 MW fusion reactor that is currently in the design phase). Furthermore, the measurements may have to be performed after the reactor is temporarily shut down. During such shut-downs, the ITER in-vessel environment will be highly radioactive due to neutron activation of the reactor wall (gamma radiation field of 3xl04 Gy/hr), the vessel will be maintained under ultra-clean vacuum conditions (<10-5 Pa), and the plasma facing surfaces will be maintained at elevated temperature (200°C). Therefore, the precision measurements must be conducted remotely, and because of the large dimensions of the fusion reactor and the limited number of access ports, the measurements must be done from large distances. The reactor vessel is a large toroid. Fig. 1 shows the cross section of the toroidal vacuum vessel, and the location of the metrology probe. Because of the very large surface area involved, fast scanning capability is also needed to reduce the reactor down-time."

Jan 1, 1997

By Pinakin Chaubal

The campaign life of a modern blast furnace depends on hearth wear. Distributions of liquid iron flow and refractory temperatures have a significant influence on hearth wear. A 3-D comprehensive CFD model has been developed specifically for simulating the blast furnace hearth. It includes both the hot metal flow and conjugate heat transfer thorough the refractories. The model has been extensively validated using measurement data from actual blast furnaces. Good agreements between measured and calculated refractory temperature profiles haven been achieved. It has been used to analyze the velocity and temperature distributions and wear patterns at different furnaces and conditions. The results can be used to point out the directions for protecting the hearth.

Jan 1, 2006

By M. Cowling, S. Kumarl, A. W. Piper, R. A. Forsdick, C. Bailei, M. Patell

"The lead-refining process takes place in hemispherical vessels called kettles. In these kettles, lead-bullion is mixed at a certain temperature to remove impurities from the melt. This paper presents steps in a novel method for calculating vortex shape and flow field in the refining kettle using Computational Fluid Dynamics (CFD). First of all a two-dimensional axisymmetric model is used to calculate the vortex shape at the liquid-air interface. Here action of impellers and baffles are accounted for by adding source and sink terms to the appropriate momentum equations. The model includes Yolume of Fluid (YOF) method to predict the shape of interface deformation, i.e. vortex. This strategy reduces the computationally intensive three-dimensional transient simulations to a more efficient two-dimensional model. The calculated vortex shape is then used to regenerate a full three-dimensional grid. This grid is used to carry out steady-state simulation using the standard multiple-reference-frame method of impeller-fluid-baffle interaction within the commercial code Fluent. Simulation results for both stages of the development are compared with experimental data.IntroductionOne of the most important steps in Lead Refining process is impeller stirred mixing of Lead Bullion in hemispherical vessels. During this process, the melt is treated with one or more reagents in order to remove impurity elements such as copper, antimony, silver and bismuth. These impurities react selectively with the reagents and float to the melt surface in the form of a powdery layer, which is then skimmed for further processing. Figure-! shows (a) the lead refining kettle and (b) the schematic representation of involved flow pattern. Stirring of the lead bath using specific impeller designs, which creates a central vortex, provides a mechanism for draw-down and dispersion of the buoyant reactants. Clearly, effective control of the impeller generated vortex and flow pattern is crucial to achieving efficient operating conditions."

Jan 1, 2001

By R. P. King

The central idea presented in this paper is straight forward. Some fairly obscure mathematical techniques are presented from the fields of integral geometry and stereology and these techniques are shown to provide workable quantitative relationships from which the liberation spectrum of an ore after comminution can be determined. The experimental data that is required can be obtained very accurately and conveniently by image analysis of polished sections of the original ore. This quantitative mineralogical analysis provides a bridge that can link the worlds of the mineralogist and the mineral process engineer. This paper has been written to introduce this quantitative approach to a wider audience in the hope that the approach will be used more widely and that the basic linear intercept analysis data will be measured for a wider range of mineralogical textures. This could lead to a more comprehensive characterization of typical mineral ores that will be of considerable use in practical mineral processing.

Jan 1, 1991

By H. Cepin

Production of non-ferrous metals is still principally performed in shaft and flame fired furnaces. Electric furnaces increase metal recovery, reduce environmental burdens and decrease energy consumption compared to conventional processes. Because non-ferrous metals have low melting and boiling points and aggressive slags design of the furnace, energy input, and slag conductivity and composition are very important. Secondary materials are easily handled in electric furnaces. Since additional amounts of lead, zinc and tin will become available from secondary sources in the future, electric furnaces are expected to replace conventional smelting furnaces. The paper describes technical, electrical, and process designs of electric furnaces peculiar to the reduction of non-ferrous metals. Electric and various thermic furnace processes are compared.

Jan 1, 1990

By Richard Jack De Klerk

The standard industrial method of arsenic removal from acidic aqueous processing effluents involves coprecipitation of arsenic with ferric iron (Fe/As>3) by lime neutralization in continuous circuits. To better understand current industrial practice, continuous circuit coprecipitation experiments were performed to identify industrially relevant process parameters and co-ion solutes that contribute to improved arsenic removal and retention. The coprecipitated solids were subjected to accelerated ageing at 20°, 40° & 70°C and arsenic behavior was monitored and modeled. Staging and the presence of calcium ions were found to have the most significant impact on arsenic retention. In this paper the techniques used and major results obtained are discussed.

Jan 1, 2008

By A. A. Hejja

"Liquidus temperature and electrical conductivity data measured on synthetic slags were modelled by the use of neural nets (NN). In this work the applied multilayer feedforward NNs were trained by a conjugate-gradient optimisation for which a three layer formulation was used. Very good fits were obtained for both the liquidus temperature and the conductivity data. The synthetic slags were prepared from pure oxides to represent a wide range of compositions likely to be encountered in ferromanganese and silicomanganese smelting. The slag constituents were in the following range: Mn0;5-30%, Ca0;20-35%, Mg0;5-15%, Si02;27-58%, Al20 3;5%. Liquidus temperatures varied from 1300° to 1380° and increased with increasing basicity ratio. The electrical resistivity of slags decreased with the increase of basicity ratio from 0.55 to 1.1 but above 1.1 basicity ratio the resistivity tended to increase depending upon the MnO content.Neural NetsVarious neural net (NN) topologies such as Hopfied nets, Hamming nets, Carpenter/Grossberg classifiers have been discussed in some detail by Lipmann (1). In this work, the applied multilayer feedforward NNs are not trained by backpropagation (BP) using the generalized delta rule (2) (GDR), but by a conjugate-gradient optimisation [3,4] - a procedure similar to that proposed by Barnard and Cole [5]. A neural net is capable ·of learning the non-linear functional relationship existing between a set of inputs and outputs [ 6]. In this capacity a neural net is used to determine the functional relationship that exists . between N dependent and independent variables, which characterize the system under consideration For a three layer neural net [1] the following function may be formulated, which relates X; inputs to the output Y 0 from node o:"

Jan 1, 1997