Flow separation in a forced oscillation glottal model

To determine the causes and history of atherosclerosis it is necessary to understand the hemodynamic parameters of blood circulation. Hemodynamic parameters play an important role in the formation of atherosclerotic plaques, especially near bends and bifurcations where the flow separates from the wall. Here the flow is laminar and non-axial with eddies, secondary flow, flow separation and stagnation points. Stenoses are found predominantly in flow separation areas. Therefore, it is important to separately study the following flow parameters: steady and pulsatile flow, wall elasticity and non-Newtonian flow behavior of blood. A simplified silicon elastic y-model simulating the human carotid artery was used for the analysis of these parameters. This model can be used for numerical studies as well. Flow was visualized at steady flow using dyes and at pulsatile flow with a photoelastic apparatus and a birefringent solution. The local axial velocity at steady and pulsatile flow was determined with a one-component Laser-Doppler-Anemometer (LDA). Pulsatile flow was generated by a piston membrane pump. A glycerin-water solution was used to simulate the Newtonian flow behavior of blood. A DMSO-Separan water solution was used to simulate the non-Newtonian flow behavior. Pulsatile flow creates higher and lower shear rates so called oscillating shear rate compare to steady flow depending on the velocity amplitude. The non-Newtonian fluid showed a markedly different flow behavior than the Newtonian fluid especially in areas of flow separation. Shear gradients were calculated from these velocity measurements using a bicubic spline interpolation. Shear stresses were calculated from these velocity shear gradients and the viscosity of the non-Newtonian fluid at these shear gradients. At special areas, high shear stresses > 10 Pa were found. The elasticity of the model wall also influences the flow behavior. The measurements showed that the characteristics of pulsatile flow and the elasticity of the model wall should be observed concomitantly. This paper presents the steady and pulsatile flow with a Newtonian and non-Newtonian fluid in an elastic model.

Airflow patterns of a hemilarynx were simulated numerically by the method of computational fluid dynamics (CFD) and were investigated experimentally with a flow visualization method. In the numerical model, vocal fold motion was simulated by sinusoidal variations of the inferior and superior glottal widths at different phase conditions. The straight glottal wall alternated between converging and diverging shape, while the input airflow varied sinusoidally between zero and a maximum. The control variables were oscillatory frequency, Reynolds number, and phase difference of the inferior and superior sections used for glottal motion. Flow visualization was performed using a hemilarynx physical model with a rectangular-shaped glottis. Using a Kodak high-speed camera, airflow patterns were recorded for various glottal widths and flow rates. CFD simulation described the complete airflow pattern and velocity distribution which indicated the existence of a high-speed jet near the fixed wall, flow separation, and vortex formations behind the glottis. Flow visualization also revealed similar patterns of flow separation, near-wall jet, and vortices below the jet. [Work supported by NIDCD Grant No. R01 DC03566.]

In order to develop a wall function boundary condition for high-speed flows so as to reduce the grid-dependence of the simulation for the skin friction and heat flux, a research was performed to improve the compressible wall function boundary condition proposed by Nichols. Values of parameters in the velocity law-of-the-wall were revised according to numerical experiments and the expression of temperature law-of-the-wall was modified based on theoretical analysis and numerical simulation. Besides, the formula of the heat conduction term in near-wall region was derived so that the coupling between the wall function boundary condition and CFD code was realized more accurately. Whereafter, the application study of the modified wall function was carried out. The numerical case of supersonic turbulent boundary layer on a flat plate illustrated that the modified wall function produces reasonable results of skin friction and heat flux, and profiles of velocity, temperature and turbulent eddy viscosity for coarse grids with the initial wall spacing of y+<400, and that the modifications to the original wall function can obviously improve the simulation precision. As for the application of separation flows, it was found from the numerical cases of supersonic cavity flow and hypersonic axisymmetric compression corner that the compressible velocity law-of-the-wall originally established based on the fully-developed attached turbulent boundary layer approximately holds in the near-wall region inside the separation flows, which ensures that reliable skin friction and heat flux can be given by the wall function inside the separation flows, while for the region near separation and reattachment points, the wall function gives results with a relatively large error, because the velocity law-of-the-wall used in the wall function takes on obvious deviation from the real velocity profiles near the separation and reattachment points.

Experimental and numerical modeling of the gas atomization nozzle for gas flow behavior

The development of accurate and efficient numerical methods for simulating incompressible flows is a challenging area of computational fluid dynamics because of its importance in a wide range of physical and industrial applications. Most of the numerical methhods used for simulating incompressible flows are finite volume and finite difference methods. Although very good predictions have been achieved by using these methods, there are still several laminar flow problems where very fine grids have to be used for predicting the correct flow structure. Such flow problems concern nonlinear phenomena, separated flows and instabilities. The very rapid advances in the computing power provide the possibility to study complex flow problems using very fine grids, but it is not still enough for studying with high accuracy unsteady, separated three-dimensional flows. Therefore the development and investigation of new numerical methods still remains a basic research direction in computational fluid dynamics.

Conventional laser cutting is optimized for cut speed and quality but does not consider process fume. When processing hazardous materials in environments where global extraction is not practical, such as during in situ nuclear decommissioning, consideration of process fume is critically important. Fume generation in laser cutting is governed by forces applied to the molten cut front; dynamic interactions between the assist-gas and the cut front control these forces. This paper investigates the causal links between key gas-dynamic features and fume generation mechanisms in laser cutting. Its aim is to provide the preliminary understanding required to design process setups targeted at reducing the fine particle fraction. During laser cutting, the assist-gas boundary layer separates (BLS) from the molten cut front and, therefore, downstream forces applied on the melt are significantly changed. In this work, the BLS point is analyzed, captured via schlieren imaging, in an idealized environment. This separation point is related to features on the cut wall of laser-cut samples, validating the gas-dynamic interactions. The samples were cut from 3 mm thick 304 stainless steel, using a 1 kW Nd:Yb fiber laser. Control of the surface stagnation pressure and gas flow boundary conditions were used to relocate the point of BLS within the cut front, and the relationship of these parameters was analyzed. The results demonstrate that shifting the point of BLS along the cut front can be used to control fume generation mechanisms. This study successfully provided the preliminary understanding necessary to manage fume generation during the laser cutting process.

With aerospace industries wishing to take maximum advantage of computational modeling CFD (computational fluid dynamics) has found increasing use at off-design conditions. These are characterized by significant areas of separated flow and the need for the resolution of small scale complex geometrical features. This new aerospace CFD direction has increased the common links between electronics CFD needs. This link is explored. As part of this, it is demonstrated how for typical non-separated aerospace flows with traditional turbulence models (RANS) CFD yields high solution accuracy (integrated stresses to within 1%). More separated aerospace flows, and the modern eddy resolving approaches now used to solve these are then considered. Despite the relatively simple level of geometrical complexity (in relation to electronics) and aerodynamic nature of geometries a significant reduction in solution accuracy is shown. However, even so the predictive accuracy could be better than 10 %. The same model types are then applied, by the same research group, with high accuracy well validated CFD codes (again with aerospace Rolls Royce - origins) to a non-isothermal ribbed channel flow, cube flow and complex geometry CPU system flow. With increasing geometrical complexity solution accuracy is shown to rapidly deteriorate. For the most complex geometry, turbulence models are shown to give a vast range of heat transfer levels deviating from measurements by around /spl plusmn/100 %. The important question of solution uniqueness is considered (for both aerospace and electronic geometry flows) and the matter of how different solvers can give qualitatively different flow solutions even for very simple multiple boundary isothermal flows noted.

Image based Computational Fluid Dynamics (CFD) simulation of the cardiovascular system is increasingly becoming important and its application in everyday medical practice can already be envisaged. The goal of this workshop is to address all the factors involved in the development of a computational framework/software for modelling and analyses of the cardiovascular system and provide examples. The development of such framework, requires integration, management and interpretation of data from several technology areas such as a) feature detection and extraction of arterial geometry from imaging data, b) adaptive grid generation techniques for 3-D asymmetric geometries c) hemodynamic modelling, disparate length-scale model, and fluid-tissue interaction with high-performance computing, d) CFD data validation, e) feature extraction/detection and visualisation algorithms, f) graphical user interface to allow remote visualisation of post processed data. These computational tools are employed to study flow in specific problem sites in the vascular tree such as the carotid, femoral, coronary and abdominal arteries. Such studies provide understanding of the factors involved in the initialisation and evolution of arterial disease due to altered flow conditions (as a result of plaque formation) such as flow separation and reversal, and low and oscillatory wall shear stress. It is also used to study the effect of various clinical procedures such as the implantation of stents, vascular grafts, vascular prostheses and artificial valve implants on local and global hemodynamics. This workshop will address a new emerging paradigm in clinical practice known as predictive medicine for effective surgical planning and post surgical rehabilitation. The workshop will also address the difficulties in the implementation of some of the technology areas in this application with examples of carotid, femoral, and abdominal artery simulations.

Chapter 14 - Parallel Personal Computer Applications of CFD++, a Unified-Grid, Unified Physics Framework

High-Resolution Modeling of Aqueous Humor Dynamics in the Conventional Outflow Pathway of a Normal Human Donor Eye

Introduction: A delayed detached eddy simulation in Open FOAM was performed to study flow separation of a circular cylinder with roundness imperfection up to 4% of its diameter at Reynolds numbers of 100, 3900, and 104 in normal flow. Methods: The flow was considered to be Newtonian and incompressible. The separation position was determined independently based on surface pressure distribution and skin friction. Results: Results show that the patterns of these distributions depend on both Reynolds number and roundness imperfection level, and flow separation in an imperfectly round cylinder may be induced by either an adverse pressure gradient or a Gentle Bend (GB) introduced by the roughness. For the separation point determined by the pressure-based method, its accuracy can be affected by the characteristic of pressure distribution near the separation point at low Reynolds numbers, and, thus, its physical validity needs to be verified by flow visualization at high Reynolds numbers. Conclusion: The skin friction-based method can accurately predict separation point for both perfectly and imperfectly round cylinders without additional information. When the roundness imperfection ratio reaches 2% and the Reynolds number reaches 3900, both approaches indicate that the flow separation point converges to the location of GB on the cylinder surface and the two sets of predicted separation points agree well.

DNS for flow separation control around an airfoil by pulsed jets

Occurrence of flow separation within the boundary-layer is primarily attributed to the growth of adverse pres-sure gradient within the boundary-layer. Such a separating boundary-layer can, however, be made to stay attached to the bounding surface due to various remedies where the surface transverse curvature (TVC) is one of them while the body contour is another. Assistive role of the surface transverse curvature is utilized in delaying the boundary-layer separation in a retarded flow past the slender bodies by assuming different body configurations. Axially-symmetric bodies of revolution of different shapes having a varying cross section in x, following the power-law form, have been considered in this study. It is shown that the point of separation is strongly dependent upon the surface curvature parameter k and the body-shape index n, which in fact control the flow separation. This fact ensures that an increase in the surface curvature parameter k and the body-shape index n leads to an increase in the wall shear stress which consequently delays the flow separation. Significant delay in flow separation is observed on cylinders having sufficiently large transverse curvature. Percent increase in the separation length corresponding to various values of the curvature parameter and the body-shape index has been calculated and reported in the form of Tables.

A computational assessment or even quantification of shear induced hemolysis in the predesign phase of artificial organs (e.g., cardiac assist devices) would largely decrease efforts and costs of design and development. In this article, a general approach of hemolysis analysis by means of computational fluid dynamics (CFD) is discussed. A validated computational model of a microaxial blood pump is used for detailed analysis of shear stress distribution. Several methods are presented that allow for a qualitative assessment of shear stress distribution and related exposure times using a Lagrangian approach and mass distribution in combination with shear stress analysis. The results show that CFD offers a convenient tool for the general assessment of shear-induced hemolysis. The determination of critical regions and an estimation of the amount of blood subject to potential damage in relation to the total mass flow are shown to be feasible. However, awareness of limitations and potential flaws in CFD based hemolysis assessments is crucial.

In this study, a fully coupled fluid-structure interaction (FSI) software system for a pulsatile flow across a moving tilting-disc valve with clearance is developed. Unsteady pulsatile flow coupled with induced valve motion has been examined in details. State-of-the-art computational fluid dynamics (CFD) methods are adopted in the present flow solver development. To account for nonstationary valve motion, the meshes surrounding the valve are generated and updated in each time-marching step using hybrid grid method. A single-degree-of-freedom rotational valve model is integrated simultaneously with the CFD adaptive time-stepping. It is found in the present study that, on both side of the occluder, strong shedding vortices occur and persist in the valve closing phase. These closure vortices show great influence on the prediction of the regurgitate flow characteristics and the subsequent valve opening dynamics as well. Based on the present software system, the results obtained from quasi-steady simulations performed at various instants of interest with prescribed valve motion are critically evaluated to assess whether simplified flow and valve conditions may lead to erroneous conclusions.