Computation of the airflow in a pilot scale clean room using K- turbulence models

This work deals with the assessment of the airflow in a food-processing clean room. The flow pattern inside the working area of a pilot scale clean room was numerically investigated using a computational fluid dynamics code based on a finite volume formulation. Two versions of the k- turbulence model were tested: the standard and the RNG version. The analysis of the velocity magnitude does not reveal sensitive differences between them. Moreover, both models well predict the main features of the flow and numerical results agree with experimental measurements. However, a further examination shows that the RNG k- turbulence model predicts more swirls and more complex trajectories. As the standard k- model overestimates the turbulent diffusion, the RNG version seems to be more suitable to calculate the airflow in clean rooms. The influence of initial turbulence intensity is also pointed out. Finally, the study of the airflow below a laminar flow unit confirms that the design of clean rooms can benefit from the numerical approach.     

Surface tension and viscosity from damped free oscillations of viscous droplets

Damped oscillations of a viscous droplet in vacuum or in an inert gas of negligible density are considered. The dependence of the complex decay factor on the properties of the liquid is investigated for the first time, and numerical results are compared with earlier studies for special cases. A new method is developed to determine both surface tension and viscosity from a single experiment in which the damping rate and frequency of oscillations are measured. The procedure to determine surface tension and viscosity from oscillating levitated liquids is outlined, and results are presented for various modes of shape oscillations.     

Simulation of an orifice scrubber performance based on Eulerian/Lagrangian method

A mathematical model based on Eulerian/Lagrangian method has been developed to predict particle collection efficiency from a gas stream in an orifice scrubber. This model takes into account Eulerian approach for particle dispersion, Lagrangian approach for droplet movement and particle-source-in-cell (PSI-CELL) model for calculating droplet concentration distribution. In order to compute fluid velocity profiles, the normal k− turbulent flow model with inclusion of body force due to drag force between fluid and droplets has been used. Experimental data of Taheri et al. [J. Air Pollut. Control Assoc. 23 (11) (1973) 963] have been used to test the results of the mathematical model. The results from the model are in good agreement with the experimental data. After validating the model the effect of operating parameters such as liquid to gas flow rate ratio, gas velocity at orifice opening, and particle diameter were obtained on the collection efficiency.     

Mathematical simulation of electromagnetic stirring of liquid steel in a DC arc furnace

Results are given of numerical simulation of electromagnetic stirring of metal melt in a dc arc furnace. The flow pattern and the transport of passive admixture in baths with one and two electrodes are studied. The mathematical model describes three-dimensional turbulent flow of electrically conducting liquid in the field of gravitational and electromagnetic forces. The parameters of turbulence are calculated in two approximations, namely, unsteady-state approximation by the large eddy simulation (LES) model and quasisteady-state approximation by the k-ε model.     

Calculation of characteristics of turbulent poiseuille flow

A previously described approach is used to calculate decay of a laminar flow into individual turbulized liquid layers. The minimum turbulence scale, turbulent viscosity, and frequency spectrum are determined.Translated from Inzhenerno-Fizicheskii Zhurnal, Vol. 57, No. 3, pp. 382–386, September, 1989.     

High-speed imaging and CFD simulations of a deforming liquid metal droplet in an electromagnetic levitation experiment

Electromagnetic levitation of a liquid metal droplet is of great interest to study gas–liquid metal reactions. An important prerequisite for the evaluation of the overall mass transfer between the gas and metal is to characterize the geometry of the deforming molten droplet, which determines the interfacial reaction area. In this article, the free surface shape and dynamics of a molten 80%Ni–20%Cr droplet is investigated both experimentally and numerically. The frequencies associated to the oscillatory translational motions of the drop and to the vibrations of its free surface are measured using high-speed video image analysis. A 2D transient model is then presented, in which three interacting phenomena are considered: electromagnetic phenomena, the turbulent flow of liquid metal in the drop and the change in the drop shape. The numerical results presented demonstrate the capabilities of the model.     

基于SPH方法的低韦伯数下三维液滴碰撞聚合与反弹数值模拟研究

光滑粒子流体动力学方法（SPH方法）作为纯拉格朗日粒子方法,可以有效避免网格法在模拟大变形过程中带来的网格扭曲等缺陷,适合模拟含大变形的液滴碰撞聚合与反弹过程。该文基于Ott和Schnetter提出的修正SPH方法,利用有限差分与SPH一阶导数相结合的方法处理粘性项中的二阶导数问题,进行Couette流算例验证,数值解...     

Numerical Simulation of Unsteady Flows and Shape Oscillations in Liquid Droplets Induced by Deployment Needle Retraction

Retractable opposed needles are often used in reduced-gravity droplet combustion experiments to deploy droplets prior to ignition. Needle retraction induces droplet shape oscillations and internal flows that can have important effects on subsequent droplet behaviors. In the present paper, the unsteady flows and droplet shape oscillations associated with deployment needle retraction are investigated using the commercial CFD software package Fluent. A volume-of-fluid method with a second-order upwind scheme and a dual time stepping solver is employed to solve the conservation equations in 2-d and 3-d simulations of droplets prior to ignition. A moving-mesh method is employed to model needle movements. Calculations indicate that rapid needle retraction causes ligament formation between a droplet and a needle, with ligament breakage sometimes resulting in the formation of satellite droplets. Needle retraction also induces droplet shape oscillations that rapidly decay, though significant internal flows are predicted to remain within droplets even after droplet shape oscillations have damped to low levels. The calculations indicate that the initial needle spacing can have important effects on droplet shape oscillations and internal flow characteristics. Comparison of model predictions and experimental data is favorable.     

Intermittent Switching Between Potential Flow and Turbulence in Superfluid Helium at mK Temperatures

Superfluid flow around an oscillating microsphere is investigated at temperatures down to 25 mK. Stable laminar flow below a critical velocity and turbulence at large drives are found to be separated below 0.5 K by an intermediate range of driving forces where the flow is unstable, intermittently switching between laminar and turbulent phases. We have recorded time series of this switching phenomenon and have made a statistical analysis of the switching probability. The mean lifetime of the turbulent phases grows with increasing drive and becomes infinite at a critical value. Stability of the laminar phases above the critical velocity is limited by natural background radioactivity or cosmic rays.     

Physical property measurements for the mathematical modeling of fluid flow in solidification processes

Measurement methods are being developed to provide values for the density, viscosity, heat capacity, enthalpy, fraction solid, surface tension, and thermal diffusivity and conductivity of commercial alloys in the liquid and mushy states. These data are needed for the mathematical modeling of heat and fluid flow in solidification processes. This paper briefly describes the present state of development of apparatus for the measurement of density by the levitated drop and hydrostatic probe methods and viscosity by the oscillating viscometer in our laboratory. Paper presented at the Fourth Asian Thermophysical Properties Conference, September 5–8, 1995, Tokyo, Japan.     

Melt Flow Instability and Turbulence in Electro-Magnetically Levitated Droplets

This article addresses fluid flow instabilities and flow transition to turbulent chaotic motions through numerical analysis and turbulence in electro-magnetically levitated droplets through direct numerical simulations. Numerical implementation and computed results are presented for flow instability and turbulence flows in magnetically levitated droplets under terrestrial and microgravity conditions. The linear melt flow stability is based on the solution of the Orr-Sommerfeld linearized equations with the base flows obtained numerically using high order numerical schemes. The resulting eigenvalue problems are solved using the linear transformation or Arnold's method. Melt flow instability in a free droplet is different from that bounded by solid walls and flow transits to an unstable motion at a smaller Reynolds number and at a higher wave number in a free droplet. Also, flow instability depends strongly on the base flow structure. Numerical experiments suggest that the transition to the unstable region becomes easier or occurs at a smaller Reynolds number when the flow structures change from two loops to four loops, both of which are found in typical levitation systems used for micro-gravity applications. Direct numerical simulations (DNS) are carried out for an electro-magnetically levitated droplet in a low to mild turbulence regime. The DNS results indicate that both turbulent kinetic energy and dissipations attain finite values along the free surface, which can be used to derive necessary boundary conditions for calculations employing engineering k--ε models.     

Numerical simulation of cryogenic cavitating flow by an extended transport-based cavitation model with thermal effects

Thermodynamic effects on cryogenic cavitating flow is important to the accuracy of numerical simulations mainly because cryogenic fluids are thermo-sensitive, and the vapour saturation pressure is strongly dependent on the local temperature. The present study analyses the thermal cavitating flows in liquid nitrogen around a 2D hydrofoil. Thermal effects were considered using the RNG k-ε turbulence model with a modified turbulent eddy viscosity and the mass transfer homogenous cavitation model coupled with energy equation. In the cavitation model process, the saturated vapour pressure is modified based on the Clausius-Clapron equation. The convection heat transfer approach is also considered to extend the Zwart-Gerber-Belamri model. The predicted pressure and temperature inside the cavity under cryogenic conditions show that the modified Zwart-Gerber-Belamri model is in agreement with the experimental data of Hord et al. in NASA, especially in the thermal field. The thermal effect significantly affects the cavitation dynamics during phase-change process, which could delay or suppress the occurrence and development of cavitation behaviour. Based on the modified Zwart-Gerber-Belamri model proposed in this paper, better prediction of the cryogenic cavitation is attainable.     

The Damping Coefficient of a Pressure Wave under Conditions of Pulsating Turbulent Flow of Gas in a Channel

The available results of experimental and prediction studies of the damping coefficient and phase propagation velocity of waves under conditions of pulsating turbulent flow in a narrow channel are reviewed and analyzed. It is demonstrated that the concept of complex damping coefficient may be introduced strictly on condition of certain restrictions imposed on an oscillating and time average flow. The dependences of the complex damping coefficient on the oscillation frequency and on the Reynolds and Mach numbers of time average flow are analyzed. The calculations are performed using the data on the relative amplitude and phase of oscillation of the tangential wall stress, obtained with the aid of the turbulence model including the relaxation equations for turbulent viscosity and tangential stress. It is demonstrated that quasi-stationary models of turbulence are invalid in the region of relatively high frequencies. Numerical simulation based on the difference solution of a set of channel cross section-averaged equations of motion, continuity, and energy is performed with due regard for the experimental conditions and measuring techniques. The calculation results agree well with the experimental data.     

Verification of optimal models for 2D-full loop simulation of circulating fluidized bed

To verify the optimal models for a two-dimensional (2D) full loop simulation of a circulating fluidized bed (CFB), different turbulent models and drag models are studied according to relevant pressure profile, voidage distribution and particle collision energy. With regard to a laminar model and turbulent models including Standard k-ε, RNG k-ε and Realizable k-ε, the experimental data reveals that the RNG k-ε model is the best at predicting pressure, voidage, axial solid velocity and granular temperature. Besides, through the comparison of four drag models, it is found that the Gidaspow model can achieve a higher accuracy of prediction. Therefore, it can be concluded that the combination of the RNG and Gidaspaw models is suitable for the 2D full loop simulation of a CFB, and therefore potential models for the prediction of flow characteristics.     

Measurements in the laminar and turbulent regime of superfluid4He by means of an oscillating sphere

The translational oscillations of a sphere in liquid helium have been measured as a way of studying superfluid turbulence. Experiments were carried out in the laminar flow regime for reference purposes, and good agreement found between measured and calculated quantities. In the turbulent region, the dissipation is found to be proportional to the square of the velocity of the sphere, as found previously by other workers. For high vibration amplitudes there is an increase in the hydrodynamic mass. This seems to scale with the superfluid fraction in a way that strongly suggests that the superfluid component plays an important role in the turbulent regime.     

Dynamics of oscillating viscous droplets immersed in viscous media

Summary Damped oscillations of a viscous droplet immersed in a viscous medium are considered in detail. The characteristic equation is solved numerically for arbitrary, finite fluid properties. The cylinder functions in the characteristic equation are solved using an accurate continued fraction algorithm, and the complex decay factor is searched using a minimization scheme. Oscillation frequency and damping rate results are presented for the fundamental mode, for various cases of practical interest (liquid-gas, and liquid-liquid systems), and the effect of the external medium properties are discussed. Results are compared to exact solutions for limiting cases, and to existing experimental data for both the fundamental and higher order modes. It is shown that the theoretical frequency prediction matches well with the experimental observation. Damping rate predictions, however, underestimate experimental observation in some cases, and this is thought to be due to surface impurities. The application of these results to the measurement of surface tension and viscosity of liquid droplets from single-droplet levitation experiments is also discussed.     

Laboratory experiments on frazil ice growth in supercooled water

Presented herein are the results of a laboratory investigation of the influence of turbulence and water temperature on the nature and rate of frazil ice growth in a turbulent body of supercooled water. The results indicate that the rate and the quantity of frazil ice growth increase with both increasing turbulence intensity and with decreasing water temperature at the instant of seeding, when a small fragment of ice is placed in the supercooled water. The turbulence characteristics of a flow affect the rate of frazil-ice growth by governing the temperature to which the flow can be supercooled; by influencing heat transfer from the frazil ice to surrounding water; and by promoting secondary nucleation, crystal, platelet and floc fracture, thereby increasing the number of nucleation sites available for further frazil ice growth.Larger frazil ice platelets, beginning as single crystals then becoming laminar fusions of crystals, were observed to form in water supercooled to lower temperatures. However, platelet size generally decreased with increasing turbulence intensity, as platelets with a major diameter larger than a certain value tend to break when buffeted by turbulence eddies.The investigation of frazil ice growth involved the use of a simplified analytical model, in which the rate of frazil ice growth is related to temperature rise of a turbulent volume of water due to the release of latent heat of fusion of liquid water to ice. Experiments conducted in a turbulence jar with a heated, vertically oscillating grid served both to guide and to calibrate the analytical model as well as to afford insights into frazil ice growth.     

Surface tension of γ-TiAl-based alloys

Within the Integrated Project IMPRESS, funded by the EU, a concerted action was taken to determine the thermophysical properties of a γ-TiAl-based alloys, suitable for casting of large turbine blades for aero-engines and stationary gas turbines. The challenge was to develop a castable alloy, free of grain refiners and susceptible to heat treatment. Owing to the high reactivity of this class of alloys, many difficulties were encountered to process the liquid phase in a crucible. This prevented also the measurements of specific heat, viscosity and electrical conductivity in the liquid phase. However, surface tension and density could be measured using container-less techniques. For the surface tension determination, both the oscillating droplet method by the electromagnetic levitation as well as a combined method using two methodologies in one test (i.e. the pendant drop and sessile drop) by an advanced experimental complex that has been designed for investigations of high temperature capillarity phenomena were applied. All the quantities have been obtained as a function of temperature, in some cases also in the undercooled liquid. In this article, we report a comparative discussion on the results obtained for the surface tension of Ti–Al–Nb and Ti–Al–Ta alloys, together with the corresponding theoretical values calculated by thermodynamic models.