Applications of CFD Method to Gas Mixing Analysis in a Large-Scaled Tank

The computational fluid dynamics (CFD) modeling technique was applied to the estimation of maximum benzene concentration for the vapor space inside a large-scaled and high-level radioactive waste tank at Savannah River site (SRS). The objective of the work was to perform the calculations for the benzene mixing behavior in the vapor space of Tank 48 and its impact on the local concentration of benzene. The calculations were used to evaluate the degree to which purge air mixes with benzene evolving from the liquid surface and its ability to prevent an unacceptable concentration of benzene from forming. The analysis was focused on changing the tank operating conditions to establish internal recirculation and changing the benzene evolution rate from the liquid surface. The model used a three-dimensional momentum coupled with multi-species transport. The calculations included potential operating conditions for air inlet and exhaust flows, recirculation flow rate, and benzene evolution rate with prototypic tank geometry. The flow conditions are assumed to be fully turbulent since Reynolds numbers for typical operating conditions are in the range of 20,000 to 70,000 based on the inlet conditions of the air purge system. A standard two-equation turbulence model was used. The modeling results for the typical gas mixing problems available in the literature were compared and verified through comparisons with the test results. The benchmarking results showed that the predictions are in good agreement with the analytical solutions and literature data. Additional sensitivity calculations included a reduced benzene evolution rate, reduced air inlet and exhaust flow, and forced internal recirculation. The modeling results showed that the vapor space was fairly well mixed and that benzene concentrations were relatively low when forced recirculation and 72 cfm ventilation air through the tank boundary were imposed. For the same 72 cfm air inlet flow but without forced recirculation, the heavier benzene gas was stratified. The results demonstrated that benzene concentrations were relatively low for typical operating configurations and conditions. Detailed results and the cases considered in the calculations will be discussed here.

Airside Velocity Measurement Above the Wind-Generated Water Waves

An experimental study conducted to investigate the airside flow behavior within the crest-trough region over wind generated water waves is reported. Two-dimensional velocity field in a plane perpendicular to the surface was measured using particle image velocimetry (PIV) at wind speeds ranging from 1.5 m s−1 to 4.4 m s−1. The results show a reduction in the mean velocity magnitude when gravity waves appear on the surface. A sequence of consecutive velocity fields has shown the bursting and sweeping processes and the flow separation above the waves. The results also indicate that the flow dynamics in the crest-trough region are significantly different than that at greater heights. High level of turbulence was observed in this region which could not be predicted from the measurements at greater heights. Thus, it is concluded that the quantitative investigation of the flow in the immediate vicinity of the interface is vital for an improved understanding of the heat, mass and momentum exchange between air and water.

A New Cavitation-Induced-Momentum-Defect (CIMD) Correction Approach to Track Cavitation Events

A new unsteady cavitation event tracking model is developed for predicting vapor dynamics occurring in multi-dimensional incompressible flows. The procedure solves incompressible Navier-Stokes equations for the liquid phase with an additional vapor transport equation for the vapor phase. The model tracks regions of liquid vaporization and applies compressibility effects to compute the local variation in speed of sound using the Homogeneous Equilibrium Model (HEM) assumptions. The variation of local cell density as a function of local pressure is used to construct the source term in the vapor fraction transport equation. The novel Cavitation-Induced-Momentum-Defect (CIMD) correction methodology developed in this study serves to account for cavitation inception and collapse events as relevant momentum source terms in the liquid phase momentum equations. Effects of vapor phase accumulation and diffusion are incorporated by detailed relaxation models. A modified RNG K-ε model, including the effects of compressibility in the vapor regions, is employed for modeling turbulence effects. Turbulent kinetic energy and dissipation contributions from the vapor regions are integrated with the liquid phase turbulence using relevant source terms. Numerical simulations are carried out using a Finite Volume methodology available within the framework of commercial CFD software code Fluent v.6.2. Simulation results are in qualitative agreement with experiments for unsteady cloud cavitation behavior in planar nozzle flows. Multitude of mechanisms such as formation of vortex cavities, vapor cluster shedding and coalescence, cavity pinch off are sharply captured by the supplemented vapor transport equation. Our results concur with previously established theories concerning sheet and cloud cavitation such as the re-entrant jet motion, cavity closure and the impact of adverse pressure gradients on cavitation dynamics.

The Effect of Boundary Layer Separation on Accuracy of Flow Measurement Using the Gibson Method

The Gibson method utilizes the effect of water hammer phenomenon (hydraulic transients) in a pipeline for flow rate determination. The method consists in measuring a static pressure difference, which occurs between two cross-sections of the pipeline as a result of a temporal change of momentum from t0 to t1. This condition is induced when the water flow in the pipeline is stopped suddenly using a cut-off device. The flow rate is determined by integrating, within a proper time interval, the measured pressure difference change caused by the water hammer (inertia effect). However, several observations demonstrate that changes of pipeline geometry like diameter change, bifurcations, or direction shift by elbows may produce an effect on the computation of the flow rate. The paper focuses on this effect. Computational simulations have shown that the boundary layer separates when the flow faces sudden changes like these mentioned to above. The separation may reduce the effective cross section area of flow modifying a geometry factor involved into the computation of the flow rate. The remainder is directed to quantify the magnitude of such a factor under the influence of pipeline geometry changes. Results of numerical computations are discussed on the basis of how cross section reductions impact on the geometry factor magnitude and consequently on the mass flow rate.

Dynamic Analysis of the Joined-Wing Configuration in High-Altitude Long-Endurance Aircraft Using Fluid-Structure Interaction Model

Recent studies of the joined-wing configuration of the High Altitude Long Endurance (HALE) aircraft have been performed by analyzing the aerodynamic and structural behaviors separately. In the present work, a fluid-structure interaction (FSI) analysis is performed, where the fluid pressure on the wing, and the corresponding non-linear structural deformation, are analyzed simultaneously using a finite-element matrix which couples both fluid and structural solution vectors. An unsteady, viscous flow past the high-aspect ratio wing causes it to undergo large deflections, thus changing the domain shape at each time step. The finite element software ANSYS 11.0 is used for the structural analysis and CFX 11.0 is used for the fluid analysis. The structural mesh of the semi-monocoque joined-wing consists of finite elements to model the skin panel, ribs and spars. Appropriate mass and stress distributions are applied across the joined-wing structure [Kaloyanova et al. (2005)], which has been optimized in order to reduce global and local buckling. The fluid region is meshed with very high mesh density at the fluid-structure interface and where flow separation is predicted across the joint of the wing. The FSI module uses a sequentially-coupled finite element equation, where the main coupling matrix utilizes the direction of the normal vector defined for each pair of coincident fluid and structural element faces at the interface [ANSYS 11.0 Documentation]. The k-omega turbulence model captures the fine-scale turbulence effects in the flow. An angle of attack of 12°, at a Mach number of 0.6 [Rangarajan et al. (2003)], is used in the simulation. A 1-way FSI analysis has been performed to verify the proper transfer of loads across the fluid-structure interface. The CFX pressure results on the wing were transferred across the comparatively coarser mesh on the structural surface. A maximum deflection of 16 ft is found at the wing tip with a calculated lift coefficient of 1.35. The results have been compared with the previous study and have proven to be highly accurate. This will be taken as the first step for the 2-way simulation. The effect of a coupled 2-way FSI analysis on the HALE aircraft joined wing configuration will be shown. The structural deformation history will be presented, showing the displacement of the joined-wing, along the wing span over a period of aerodynamic loading. The fluid-structure interface meshing and the convergence at each time step, based on the quantities transferred across the interface will also be discussed.

Effects of Packing and Aspect Ratio on Mixing and Heterogeneous Catalysis in Microchannels

Many industrial chemical processes involve the mixing of two or more liquids. By reducing chemical reactors to microscale dimensions, engineers seek to take advantage of decreased diffusion lengths, leading to increased effectiveness (e.g., higher purity of product) over larger process components. In this study, computational models developed using the commercial multiphysics code CFD-ACE+ are used to predict flow within microreactor channels. Two aqueous streams enter a channel—one containing a contaminant and the other devoid of the contaminant. Changes in two geometric attributes are investigated with respect to their effect on mixing of the streams: 1) packing feature layout within the channel and 2) channel aspect ratio. Reynolds numbers (Re) for the simulations range between 0.1 and 100. Results indicate that both packing feature position within the channel and channel aspect ratio can have a substantial impact on mixing. Between Re = 0.1 and Re = 1, mixing efficiency generally decreases with increasing Re; however, as the Re is increased from 1 to 100, fluid flow patterns in the channel are altered, and wake regions and streamline changes created by the packing features lead to improved mixing. Examples showing enhanced chemical conversion during heterogeneous catalysis as a result of better mixing are also presented.

Structure of Turbulence Around the Trailing Edge of a Rotor Blade

This paper focuses on the structure of turbulence around the trailing edge of a rotor blade operating behind a row of Inlet Guide Vanes (IGVs) located upstream of the rotor. High resolution, two-dimensional Particle Image Velocimetry (PIV) measurements are conducted in a refractive index matched turbomachinery facility that provides unobstructed view of the entire flow field. We focus on a small region around the rotor blade trailing edge, extending from 0.04c upstream of the trailing edge to about 0.1c downstream of it, c being the blade chord length. We examine the phase dependent distribution of turbulent kinetic energy (TKE) and its in-plane components of production rate. Impingement of an IGV wake on the suction surface of a rotor blade, near the trailing edge region, reduces the thickness of the boundary layer within the region impinged by the wake. The resulting increase in phase averaged shear strain rate increases the production rate and causes a striking increase in peak turbulent kinetic energy in the near wake. Streamwise velocity gradients, i.e. compression, also contribute to turbulence production, especially when the boundary layer at trailing edge is relatively thick, i.e. when it is not impinged by the IGV wake.

Experimental and Computational Analyses of Pressure Differentials in Flexible Ducts With Different Bent Angles

A set of experimental analyses was conducted to determine static pressure drops inside non-metallic flexible, spiral wire helix core ducts, with different bent angles. In addition, Computational Fluid Dynamics (CFD) solutions were performed and verified by comparing them to the experimental data. The CFD computations were carried out to produce more systematic pressure drop information through these complex-geometry ducts. The experimental setup was constructed according to ASHRAE Standard 120-1999. Five different bent angles (0, 30, 45, 60, and 90 degrees) were tested at relatively low flow rates (11 to 89 CFM). Also, two different bent radii and duct lengths were tested to study flexible duct geometrical effects on static pressure drops. FLUENT 6.2, using RANS based two equations - RNG k-ε model, was used for the CFD analyses. The experimental and CFD results showed that larger bent angles produced larger static pressure drops in the flexible ducts. CFD analysis data were found to be in relatively good agreement with the experimental results for all bent angle cases. However, the deviations became slightly larger at higher velocity regimes and at the longer test sections. Overall, static pressure drop for longer length cases were approximately 0.01in.H2O higher when compared to shorter cases because of the increase in resistance to the flow. Also, the CFD simulations captured more pronounced static pressure drops that were produced along the sharper turns. The stronger secondary flows, which resulted from higher and lower static pressure distributions in the outer and inner surfaces, respectively, contributed to these higher pressure drops.

Evolution of a Cavitating Bubble in a Nonuniform Pressure Field

The cavitation erosion problem is not a new one; however, it is still important and even becomes more pressing. This is associated with requirements to justify and extend service life of power-generating plants while obviously seeking the maximum efficiency. Currently semiempirical correlation relations are typically used in pump designing for prediction of modes of operation that may be hazardous in terms of development of cavitation erosion [1, 2]. It appears, however, that a progress can be made in this field by introducing numerical modeling of flow with direct modeling of the cavitation erosion process. This optimism is based on an established fact that the major effect of erosion damage is observed in the mode of operation between the “NPSH-incipient” mode (mode of activation of vapor-phase formation centers) and the “NPSH-3%” mode (mode where a noticeable vapor volume content is produced in the near-wall layer), i.e. in the mode of operation where the vapor volume content is small and its effect on flow characteristics can be neglected.

Direct Numerical Simulation of Turbulent Boundary Layer Downstream of Turbulence-Generating Grid

The effects of external grid turbulence in a free stream on the turbulent boundary layer are investigated by means of the direct numerical simulation (DNS). A square turbulence-generating grid, on which the velocity components are set to zero, is located upstream of the flat boundary with a nonslip condition. The fractional step method is used to solve the governing equations. The results show that turbulence intensities and Reynolds stress normalized by the inner parameters are significantly suppressed in the log-law region by the external grid turbulence.