Search Documents

By Lanre Oshinowo

Computational Fluid Dynamics (CFD) is used in conjunction with process modelling at HATCH to achieve a superior level of confidence in the process design of hydrometallurgical process plants. The use of these tools to evaluate designs through virtual prototyping reduces the risks associated with making design decisions. This paper highlights the development of a co-simulation strategy at Hatch for the seamless integration of process modeling software addressing different scales (process time and space): flowsheet level models and .equipment level models. With co-simulation, the process design engineers can improve design accuracy, troubleshoot and optimize performance and influence decisions impacting equipment operation by representing specific unit operations in the flowsheet model software with detailed CFD models. Specifically, this paper will focus on the co-simulation of a Ni/Co high-pressure acid leach (HPAL) process including heaters/heat exchangers and flash system, autoclave and quench/scrubber system. The knowledge developed can be transferable to other processes.

Jan 1, 2004

By K. Moriyamo

Operation improvement efforts at Hyuga Smelting Co., Ltd. have focused on calcining capacity in the reduction rotary kilns ("the kilns"), whose limited feed rate capacity had capped production at around 18,000t-NiN since 1995. To improve the kilns' feed rate capacity of 80 wet t/hr, attempts were made to increase the gas flow in the kilns by increasing the capacity of the off-gas fans. However, as the increased gas flow resulted in only limited improvement of the kilns' feed rate, a scoop feeder was installed on each kiln to add coal to the middle of each kiln's interior to gain a higher efficiency of thermal supply and conversion. With these gas flow and coal charge addition operations, the kilns' feed rate capacity increased up to 120 wet t/hr, with burner fuel efficiency increasing by approx. 10% (approx. 5% in total with the additional coal included). As a result, production was raised to a level of 22.000t-Ni/Y Keywords: reduction rotary kiln, scoop feeder, additional coal charge.

Jan 1, 2004

By J. Tang, B. Q. Li, R. Akarapu

"An integrated model is developed to represent the three-dimensional electromagnetic wave propagation and thermal phenomena during microwave heating of ceramics and dielectric/biological materials. The model development is based on the edge finite element formulation for electromagnetic fields and the node fmite element formulation for thermal phenomena. Results are presented and compared well with available analytical solutions and experimental measurements. With the model, the effect of sample positions on the electromagnetic and thermal field distributions is studied.IntroductionMicrowave, as an electro-heating source, has been widely used in the processing industries for packaged foods and other materials. The commonly used household microwave oven is a typical example of microwave heating processes. Other industrial microwave processing systems include large-scale microwave guides and cavities, which are used for the sintering operations as required for making ceramic components or for sterilizing the packaged foods."

Jan 1, 2004

By Shantaliu S. Shevade, Muhammad M. Rahman

"This paper presents a systematic analysis of fluid flow and heat transfer processes during the magnetic heating of the substrate material. Heat is generated in the substrate material when a magnetic field is imposed due to change in orientation of the molecules. As a result, the substrate temperature rises. The i;tudy considered microchannels with rectangular and square cross section with heat generation in the substrate due to imposed magnetic field. The results computed were for Gadolinium substrate and water as the working fluid. The purpose of this study is to explore the transient heat transfer coefficient when the fluid is circulated through the substrate via microchannels. Equations governing the conservation of mass, momentum, and energy were solved in the fluid region. In the solid region, heat conduction was solved. From the simulation results, plots of Nusselt number and heat transfer coefficient were obtained over the length of the channel. A thorough investigation for velocity and temperature distributions was performed by varying channel aspect ratio, Reynolds number, and heat generation rate in the channel. It was found that the peripheral average heat transfer coefficient and Nusselt number is larger near the entrance and decreases downstream because of the development of the thermal boundary layer. With the increase in Reynolds number, the outlet temperature decreased and the average heat transfer coefficient increased.IntroductionMagnetocaloric effect (MCE) is the heating or cooling of the magnetic materials due to varying magnetic field. The effect was originally discovered by Warburg. Magnetocaloric materials store heat energy in the way the atoms vibrate and in the way electrons spin within each atom. More heat energy increases the vibrations and also makes the spins more random. When a strong magnetic field is applied to a magnetic material, the magnetic moments of its atoms become aligned making the system more ordered. The more ordered material has lower entropy and compensates for the loss by heating up. When the magnetic field is removed, the magnetic moments return to their random directions, entropy increases and the material cools. Gadolinium is a magnetic material that exhibits high temperature rise during adiabatic magnetization around its transition temperature of 295 K. It has been experimentally observed that adiabatic temperature rise increases with magnetic field strength and can be as high 20 K when a magnetic field of IO T is applied. Thus magnetic field, when applied to a microchannel heat exchanger, causes temperature change in the magnetic material. The fluid flowing through the channel gets cooled or heated due to this adiabatic temperature change."

Jan 1, 2004

The high pressure acid leach process (HPAL) is being commercially used in four plants around the world to treat limonitic laterites. The two major operating costs of the HPAL process are the sulphuric acid and the energy required to heat the pulp to the desired temperature (>240°C). As an alternative to acid injection and steam heating, the in-situ generation of sulphuric acid and heat, produced during the high temperature pressure oxidation of sulphur-bearing materials blended with laterites, has been investigated at the laboratory scale. Various sources of sulphur have been examined, in particular, elemental sulphur, pyrrhotite and a Cu-Ni low-grade bulk concentrate. Under the proper blend ratios, all sulphur sources tested generated sufficient acid to result in equivalent nickel and cobalt dissolutions. A significant improvement to the process economics can result from the addition of sulphur sources containing valuable by-products, such as nickellcobalt andlor precious metals (gold and PGMs). This is especially true in the latter case, when the HPAL can be operated under Platsol conditions (i.e. with the addition of 5-10g/L of chloride), to dissolve the base metals and precious metals simultaneously in the autoclave. Various examples of the application of this concept are presented, together with a brief discussion of its implication on the autoclave design and the downstream processes to recover the values from the leach solutions. Keywords: limonite, co-treatment, acid consumption, sulphur, sulphur compounds, pressure acid leach (HPAL, PAL), Platsol process

Jan 1, 2004

By Donald J. Hallbom

Limonitic nickel laterites slurries exhibit complex rheological properties including time dependency that varies from thixotropy to rheopexy. Chronic problems may occur if plant designers and operators fail to take time dependency into account. Stress decay tests indicate that laterite slurry takes roughly two minutes to approach steady state after a step change in the shear rate. As a result, most rheometer measurements are taken at a non-equilibrium state, making evaluation difficult. Two minutes is also long relative to the time that a unit of slurry in a processing plant is at any given shear state, so the slurry may never reach equilibrium. This paper presents a method to describe the rheology of laterite slurry using a Flow Array, which describes the rheology using an array of instantaneous flow curves crossed by an equilibrium flow curve. A rate equation for structural change in non-equilibrium conditions allows time dependence to be efficiently modeled. This flow array allows the flow behavior of time- dependent slurry to be predicted in non-equilibrium conditions using a simple time-step methodology. Keywords: Rheology. nickel laterite, hydraulic transport, pressure acid leach, flow array

Jan 1, 2004

By Terry Gerritsen

Electrode control systems regulate power by physically positioning electrodes. Due to the size and weight of the electrodes, this is a slow process where power fluctuations of less than a few seconds are not fully corrected. Nickel laterite AC furnaces operating in high voltage, shielded arc mode typically incur frequent power variations of approximately +I- 20% around the power set point. To provide enhanced power control, Hatch has recently supplied, at commercial scale, a thyristor-switched reactance system called an SPLC. The SPLC incorporates predictive control software to operate furnaces at the maximum transformer rating by reducing the large power swings experienced without the SPLC. The resulting higher average power enables higher production, without increasing transformer or power plant capacity. A 60 MW production version of the SPLC has been commissioned on Falconbridge Dominicans's laterite nickel furnace, yielding excellent results with a 3 to I reduction in power fluctuations. Keywords: SPLC, AC Power Supply, Electric Furnace

Jan 1, 2004

By Mario H. Castro-Cedeno

"A design for a handheld MEMS chemical detector is proposed. The detector will be used to detect trace amounts of chemicals in air. It can detect drugs, toxic chemicals and/or explosives. The proposed detector will use ion mobility spectrometry (IMS), a technique that is well known for accuracy and has been used in the past in suitcase sized chemical detectors. The separation and detection of the chemicals occurs in the drift channel and the process was modeled using the finite element method and CoventorWare modeling software. A manufacturing process sequence for the drift channel is proposed and was demonstrated.IntroductionThe goal of this effort is to design, mode~ and prototype a MEMS chip that will be the key component in an ion mobility spectrometer (IMS). This technique uses and asymmetric potential to separate the various ion species in a stream of incoming air [1-4]. One advantage of IMS is its ability to separate chemical species from air at atmospheric pressure and temperature. This is a requirement for a simple to use handheld drug or chemical detector. IMS can be used to detect drugs, toxic chemicals and explosives and it is well known for accuracy in detecting very small amounts of chemicals. This chemical detection technique has been used in the past, but not in handheld detectors.One variant of IMS is the high-field asymmetric waveform ion mobility spectrometry or FAIMS [3]. The FAIMS process, shown in Figure l, works as follows: a gas sample is admitted into the instrument where gas molecules are ionized, typically by a weak radioactive isotope source, ultraviolet light source; and more recently through field emission. The ions move along the length of a channel while they are separated in the transverse direction by an asymmetric electrical potential. The potential, in the form of an asymmetric waveform, causes relative motion between the various chemical species. The species that hit the charged top and bottom plates are neutralized and only those ions that move along the center are detectable as charged chemical ions at the end of the channel."

Jan 1, 2004

By E. Legault-Seguin

The current industrial practice in the processing of nickeliferous laterites involves the precipitation and releaching of Ni and Co to allow separation from the major impurities of Mn, Mg and Ca. A direct solvent extraction process has significant advantages over this practice, which suffers from difficult and costly precipitation steps, lower recovery and inherently complex flowsheets. The present investigation involves the examination of Cyanex 272, Cyanex 301 and Cyanex 302 as extractants for the direct extraction of Ni and Co from major impurities associated with laterite leaching, i.e. Mn, Mg and Ca. For Co extraction, Cyanex 301 and Cyanex 302 are superior to Cyanex 272, because they are able to selectively separate Co from Mn, Mg and Ca. Cyanex 272 offers complete extraction of Co from Ni and Ca, but Mn and Mg pose difficulty. Ni can only be extracted selectively with Cyanex 301, thus another extractant selective for Ni, such as Versatic 10 or Cyanex 301, is needed to recover Ni if Cyanex 302 or Cyanex 272 is used for the initial extraction.

Jan 1, 2004

By Fan Jiang, Haigang Wang, Shi Liu

"This paper describes several applications of Electrical Capacitance Tomography (ECT) in gassolid two phase flows. A new image reconstruction algorithm was evaluated by both simulation and experiments. It is then applied to the online measurement of solids distributions in square circulating fluidized beds (CFBs) and good results were obtained. From the measurement of solids distributions in the bottom zone of a CFB, a bubbly fluidization pattern is identified. A cross-correlation technique was combined with the solids concentration measurements to obtain the flow rate through the dip leg of the cyclone, i.e. the circulating rate of a test CFB. The same· technique was applied in a dense phase pneumatic transport pipe provides the solids concentration, and the flow rate of a wavy gas-solids flow.IntroductionElectrical capacitance tomography (ECT) has been developed for imaging multiphase distributions in the 90's of the ·last century. It is a non-intrusive method. The principle of ECT can be firefly described as follows:Firstly ECT uses a number of electrodes to surround the measuring zone, depicted by Fig. I. The capacitances between the electrodes are measured in turn and an array C is obtained that contains a total number of m independently measured capacitances. Meanwhile the measuring zone is virtually divided into a grid of n pixels, represented by an array G, named the image vector, the gray level of each pixel represents the local value of the concentration of the material. As the material distribution will be reflected by the capacitances between each pair of electrodes, such distributions can be reconstructed according to the measured capacitance, a process called image reconstruction. A common image reconstruction method is called linear-backprojection (LBP). It linearize the relationship between the dielectric distributions in the measuring zone and the capacitances between the electrodes as the following:"

Jan 1, 2004

By Finlay Campbell

Nickel laterite pressure acid leach plants and reduction roast plants have had a checkered history in their engineering, construction, startup and operational reliability. Many potential projects have been the subject of detailed studies over the last thirty to forty years but relatively few have advanced to commercial production. Four reduction roast plants are currently operating, an additional one was partially constructed but never completed, and another was started up in 1974 and shut down after 12 years operation. Three acid leach plants have been built and commissioned in the last five years, trying to follow the example of the Moa Nickel plant in Cuba. An additional acid leach plant project is currently on hold, following limited construction work, its future subject to review by the owner because of potential cost over-runs. This paper attempts to discuss difficulties that may be encountered as a project is developed and will address factors which may affect production ramp-up schedule, long term production performance and overall production success. Comparisons are made with ramp-up performance at other large metallurgical plants that employ some comparable unit operations. Keywords: Caron, PAL, nickel laterite ore, ramp-up, reliability, process development, equipment design, step-out, commissioning, startup.

Jan 1, 2004

By Julian Johnson

The effect of changing redox potential (Eh) on the high temperature sulfuric acid leaching of Western Australian dry-land nickel laterite ores has been investigated. In particular, the effect of added reductant (sulfur, sodium sulfite) or oxidant (potassium chromate) on Eh, leaching kinetics, metal extractions, acid utilisation and leach residues has been determined. Three separate ores - nonhonitic ores from Murrin Mumn and Bulong, and limonitic ore from Cawse - were used in the study. Sulfur addition varied from 0 to 7.8 kg/t ore. The total acid loading was kept constant at 430 kg/t (Murrin Murrin ore) and 330 kglt (Cawse ore), whilst acid loading for Bulong ore was varied according to the presence of either reductant or oxidant. The extraction of both nickel and cobalt from limonite ore remained unaffected when small amounts of sulfkr were added, but addition of even small amounts of reductant to nontronitic ore decreased both nickel and cobalt extraction. Increased manganese and iron dissolution were observed as the levels of reductant increased, accompanied by a decrease in chromium dissolution. As the iron dissolution increased, free acidity decreased, with a subsequent decrease in nickel extraction. In contrast, addition of oxidant lead to reduced iron and manganese dissolution and slower extraction of nickel and cobalt. Keywords: Limonite, Nonhonite, Eh, Nickel Extraction

Jan 1, 2004

By Shekhar Bhansali, Shyam Aravamudhan

"This paper reports on the design and simulation results of a silicon bulk-micromachined diaphragm-type piezoresistive pressure sensor capable of operating linearly over a wide dynamic range (0-1000 meters of water). The piezoresistor elements of the sensor are arranged in a wheat stone bridge configuration to achieve a higher output response. Simulation results (output voltage and sensitivity analysis) obtained using coupled MemPZR-MemMech module of the CoventorWare2003 Finite Element tool are presented. The effect of diaphragm size, diaphragm thickness and thickness of the piezos on output sensitivity were analyzed and optimized to get a linear response over the wide dynamic range.IntroductionThe physical and biological processes in the ocean are critically influenced by ocean parameters (pressure, salinity and temperature). Thus there is a need to develop inexpensive and reliable sensors to continuously measure large volumes of sea-space over long time spans. A key component in any such system will be a pressure/depth sensor. This sensor is routinely employed in marine applications such as wave and tide gauges, tsunami detection, platform leveling etc. More significantly, pressure measurement is needed for depth determination. Depth sensors are vital for remote operated vehicles (ROVs), profiling instruments and towed arrays.Several techniques have been routinely used in the past to measure pressure in the marine environment [1-3]. In general, the pressure sensor design is based on a flexible membrane as the spring element for the sensing pressure [4]. The deflection of the membrane, induced by the applied pressure, is converted into an electrical output through a component that is sensitive to diaphragm deflection (capacitive coupling) or associated stresses (piezoresistors, strain gauges). The temperature compensated pressure measurement can be readily converted to depth using"

Jan 1, 2004

By Ernie M. Crist, Laurentiu Nastac, Frank Spadafora

"This paper describes the modeling techniques and the experimental and computer results for two casting processes that are under development and optimization. These processes are described as follows:1) Electromagnetic stirring (EMS) during casting of Ti-6Al-4V round ingots using the plasma arc cold hearth melting (PAM) process: The purpose of EMS is to establish a singlemelt PAM process capable of consistently producing as-cast ingots without helium porosity and also having no need for surface conditioning prior to secondary processing.. The modeling approach is based on the numerical solution of Maxwell's equations, fluid flow, and heat transfer equations, and mesoscopic modeling of the grain structure.2) Casting of Ti-6Al-4V slabs using the single-melt PAM process: The purpose of slab casting is to improve yield by eliminating an intermediate blooming step for converting round ingots to slabs. The modeling approach is based on the numerical solution of fluid flow and heat transfer equations. The effects of pouring and mold temperatures on the pool and temperature profiles were studied. Also, the torch effects (such as the torch standoff and its traversing path) on the pool and temperature profiles were investigated and the best-case scenario was used in the experiments. Evaluation results at RMI Titanium Company showed that the plates made from the slabs met the mechanical properties and microstructure requirements.Acknowledgment. This work was conducted by the National Center for Excellence in Metalworking Technology (NCEMT) operated by Concurrent Technologies Corporation (CTC) under contract No. N00014-00-C-0544 to the US. Navy as part of the US. Navy Manufacturing Technology Program. Dr. K. b. Yu from RMI and my colleagues from CTC are acknowledged for their comments on this work."

Jan 1, 2004

By T. Plikas, L. H. Gunnewiek, L. Oshinowo

"Electric furnace smelting is one of the principal unit operations for ferroalloy production, and increased process intensity, improved availability, minimal maintenance and a longer campaign life are common objectives for electric furnace operation in order to obtain favourable economics of production. Furnace designers have continued to develop innovative solutions, e.g., furnace cooling systems, that have allowed these objectives to be realised and CFD modelling plays a key role in the design process. Enhanced understanding of the energy transfer process in the furnace and the ability to develop, evaluate and optimize components of the crucible design to match process requirements are the key motivators for modelling.This paper provides an overview of recent applications of CFD to the design of electric furnaces. The aim is to illustrate what can be done, and identify areas where further work is needed in this field. Examples include furnace bath simulation, crucible cooling system design, taphole design, baking of Soderberg electrodes, off-gas systems, and fume control and building ventilation. It is argued that the trend is for numerical models to become more encompassing while at the same time more accessible, and therefore further improvements in furnace design will be realised with continued and well validated modelling efforts.IntroductionOne of the main processing steps in ferroalloy production is the smelting of ore in electric furnaces. The three-electrode, AC submerged arc furnace is most prevalent method for smelting in the industry, with shielded-arc and DC plasma arc furnaces achieving limited application (1,2). Commercial furnaces range in size from a few megawatts to more than eighty megawatts of supplied power, and operate with the objective of maximising the return on capital invested. Maximising productivity and energy efficiency are two pathways to achieving this objective, through process intensification, increased availability, minimal maintenance, and increased campaign life for the furnace and associated equipment. Concurrently, the need to reduce the environmental impact of metal production by controlling the emission of particulate emissions, greenhouse gases, and other harmful by-products must also be met."

Jan 1, 2004

By Harry Littleton, Jr. Scarber, Wanliang Sun

"The Lost Foam Casting (LFC) process is relatively immature when compared to other casting processes. To make the process more robust, computer modeling is needed to reduce lead times and casting defects. In this study, metal filling of lost foam aluminum castings was visualized using state-of-art real time X-Ray technology. Metal filling of the lost foam casting was simulated using a commercial software package. The result of the computer simulation was compared with the results of the real time X-Ray visualization. Some of the computer simulation results agreed with the real time X-Ray visualization, but some did not. Based on other studies using real time X-Ray technology, new input variables, such as foam density variations, were developed for the computer modeling. With the new variables, modeling of the lost foam casting process becomes more realistic. IntroductionMold filling of the Lost Foam Casting (LFC) process is an important step toward making a quality part. In the LFC process, molten metal is poured against a refractory coated foam pattern surrounded by dry sand. At the pouring temperature of aluminum, the foam pattern is decomposed into 30% (wt) gaseous and 70% (wt) liquid foam pyrolysis products. Existence of the gaseous foam decomposition products significantly increases the back pressure on the metal stream compared with the conventional open cavity sand mold. Entrapment of the liquid pyrolysis products during the metal filling generates many casting defects, such as internal pores, fold and blisters. Existence of these casting defects sometimes seriously degrades the quality of the components manufactured by the LFC process. Understanding and controlling the metal filling of the LFC process will certainly make this process more robust."

Jan 1, 2004

By Miller, Geoffrey W.

The Ravensthorpe Nickel Project (RNP) flow sheet provides for the upgrading of Ni & Co in a size based beneficiation process prior to leaching. The upgrading effect on Ni and Co makes the beneticiation process a powerful addition to the flow sheet. Multiple linear regression (MLR) analysis based on 351 individual batch tests, 52 large scale composites and 21 pilot plant tests has been used to successfully model beneticiation product grade, recovery, upgrade and mass recovery. Statistically, 83% and 96% of the variability in nickel recovery and nickel product grade respectively was accounted for. Algorithms have been developed for Ni, Co, Mg, Al and Fe, which represents the revenue and operating cost factors for the process. The algorithms have subsequently been applied to the resource model, allowing prediction of elemental grade and recovery for the development of the mine schedules based upon these parameters and not head grades alone. This has allowed optimisation of the mine schedule for improvement to the project's NPV. Keywords: Ni Laterite, Prediction, Multiple Linear Regression, Beneficiation

Jan 1, 2004

By T. T. Chen

This mineralogical study is based on samples from typical "wet" nickel laterites of humid tropical to sub-tropical regions. An in-situ nickel laterite sample from Goro and a transported laterite sample from Plaine des Lacs, New Caledonia, are predominantly limonitic. Two-thirds of the particles in the Goro sample are <10 um in size; they consist of skeletal goethite and serpentine, as well as minor amounts of talc, gibbsite, chlorite, quartz and clay minerals. The coarser grains are mainly Cr-spinel, serpentine, gibbsite and Mn oxide. Nickel occurs mainly in goethite, serpentine and Mn oxide; cobalt occurs in goethite and Mn oxide. The Plaine des Lacs sample is finer-grained and contains more Al and CO3, but less Mg and Si than the Goro material. The sample contains more kaolinite, chlorite, siderite and feldspar, but less serpentine and gibbsite than the Goro material. Nickel is detected mainly in goethite, serpentine, Fe-Mg silicate and Fe-Mn oxide. The Indonesian ore from Sorowako, as sampled from the reduction kiln feed containing some reverts, is predominantly saprolitic, with apparent particle sizes ranging from <100 um to > 1 cm. The sample of reduction kiln feed is rich in Mg silicates, Mg-Fe silicates and goethite, and exhibits complex alteration textures. The nickel contents vary from particle to particle. Nickel is present mainly in serpentine, olivine, chlorite and amphibole, and in subordinate amounts in goethite and Mn oxide. Cobalt is detected in the Mn oxide. Keywords: Nickel Laterite; Mineralogy; Nickel Carriers; Goro, New Caledonia; Plaine des Lacs, New Caledonia; PT Inco., Sorowaki, Indonesia

Jan 1, 2004

By Edwin de Jesús Deveaux

After 32 years of operations beneficiating lateritic nickel ore in a fully integrated pyrometallurgical complex, Falconbridge Dominicana (FALCONDO), situated at Bonao and La Vega Provinces, Dominican Republic, has become profitable with one of the lowest nickel grade content processing plants in the world, and literally without any by-product (granulated slag, will be utilized in the future as a post-processing subproduct). To December 2002, mineral reserves were 64.1 million tonnes containing laterite grading as low as 1.15 % Ni when full dilution is applied. At a cut-off grade of 1.20% Ni, the reserves will provide 15-20 years of production. The mineralized thickness over the seven active deposits averages 6 m. Development of very selective and unique mining practices, together with improvements in the process plant, are key to achieving the production goal (30 kt Ni/yr) and to meet the quality standards for the ferronickel produced. Keywords: Laterites, grade control, mining methods, environment

Jan 1, 2004

By C. Beckermann

"A diffuse interface model is derived for the direct simulation of two-phase flows with surface tension, density differences between the phases, and phase change. The derivation starts from the balance equations for a sharp interface and uses a formal averaging procedure on an atomic scale to obtain diffuse interface versions of the equations. Two approaches are considered: (i) the two phases have different velocities inside the interface and separate conservation equations are solved for each phase, and (ii) the two phases have the same velocity inside the diffuse interfuce and mixture conservation equations are solved. Based on a superposition of microscopic (atomic-scale) and macroscopic interface morphologies, an expression for the interfacial momentum source term is introduced that models surface tension effects and accounts for immiscibility between the phases.IntroductionDiffuse interface methods have been a popular tool in the direct simulation of two-phase flows [I]. In diffuse interface approaches, the interface between the two phases is characterized by a rapid but smooth transition in the properties, such as density and viscosity. Usually, a phase function ? is introduced to represent this transition, and the evolution of ? in the domain is computed from a separate equation or recipe. A unique set of conservation equations (e.g., mass and momentum) is solved on the entire domain, irrespective of the phase present, by accounting for the transition in the properties and introducing terms that account for interfacial sources. The continuum surface force (CSF) model of Brackbill et al. [2], which accounts for surface tension, is an example of the latter. The volume-of-fluid (VOF), level set, and phase-field methods are all examples of diffuse interface approaches. The major advantage of diffuse interface methods is that explicit tracking of the interface and the application of traditional sharp interface conditions are avoided. A disadvantage can arise from the artificial smearing of what is in reality an interface of atomic-scale width; however, techniques have been developed to overcome this problem."

Jan 1, 2004