Search results

The purpose of this study is the computational analysis of atmospheric, laminar, stoichiometric and premixed hydrogen-air flames in the presence of a quenching mesh. The assessment of the predictive capability of different reaction mechanisms, the clarification of the relative importance of the thermal and chemical effects for mesh quenching and the investigation of the influence of the mesh geometry on the quenching effectiveness are the focal points of the investigation.

The problem is posed as unsteady, two-dimensional. Differential governing equations are numerically solved by the finite volume method for the reacting hydrogen/air mixture, assuming an ideal gas behaviour. Thermal radiation and buoyancy are neglected. A coupled solver is used to treat the velocity-pressure coupling, along with a stiff-chemistry solver for the chemical kinetics. Second-order discretization schemes are used in space and time. A uniform grid resolution is used, where the grid independence in terms of the flame speed prediction is ensured in preliminary calculations for one-dimensional flames.

It is found that a detailed reaction mechanism is necessary for an accurate prediction. Meshes with round openings are found to be more effective that those with slit openings (SOs), by a factor of two in the maximum safe gap size. A perforated plate is observed to have a higher quenching potential compared to a wire mesh, for SOs. It is also found that the heat loss to the wall is the dominating quenching mechanism for the present problem, whereas adsorption of radicals plays a subordinate role.

In contrast to the previous studies in the field, a detailed reaction mechanism is applied instead of a single-step one, while still using the latter for comparison. The role of wall-radicals interaction for the quenching effectiveness of the mesh is addressed for the first time. Parametric studies are performed on the mesh geometry, which was not done before. Hydrogen is considered as fuel in contrast to the great majority of the previous work.

The purpose of this paper is the numerical investigation of the friction laws for incompressible flow in undulated channels, with emphasis on the applicability of the hydraulic diameter concept. A focal point of the study is the derivation of correlations to increase the accuracy of the hydraulic diameter approach.

Calculations are performed for laminar and turbulent flow, for Reynolds number ranges between 10–2,000 and 5,000–100,000. For turbulent flow, the shear stress transport (SST) model is used. A simple, sawtooth-like undulation shape is considered, where the channel geometry can be described by means of three length parameters. Letting each to take three values, totally 27 geometries are analyzed.

It is observed that the hydraulic diameter concept applied via analytical or empirical expressions to obtain friction coefficients does not lead to accurate results. For laminar flow, the maximum deviations of analytical values from predicted are about 70%, while 20% deviation is observed on average. For turbulent flow, deviations of Blasius correlation from predicted ones are smaller, but still remarkable with about 20% for maximum deviation and about 10% on average.

Applicability of the hydraulic diameter concept to undulated channels was not computationally explored. A further original ingredient of the work is the derivation of correlations that lead to improved accuracy in calculating the friction coefficient using hydraulic diameter. For laminar flow, the maximum and average deviations of present correlations from numerical predictions are below 5% and 2%, respectively. For turbulent flow, these numbers turn out to be approximately 12% for the maximum deviation and about 2% for the average.

The main purpose of the paper is the validation of a broad range of RANS turbulence models, for the prediction of flow and heat transfer, for a broad range of boundary conditions and geometrical configurations, for this class of problems.

Two‐ and three‐dimensional computations are performed using a general‐purpose CFD code based on a finite volume method and a pressure‐correction formulation. Special attention is paid to achieve a high numerical accuracy by applying second order discretization schemes and stringent convergence criteria, as well as performing sensitivity studies with respect to the grid resolution, computational domain size and boundary conditions. Results are assessed by comparing the predictions with the measurements available in the literature.

A rather unsatisfactory performance of the Reynolds stress model is observed, in general, although the contrary has been expected in this rotating flow, exhibiting a predominantly non‐isotropic turbulence structure. The best overall agreement with the experiments is obtained by the k‐ω model, where the SST model is also observed to provide a quite good performance, which is close to that of the k‐ω model, for most of the investigated cases.

To date, computational investigation of turbulent jet impinging on to “rotating” disk has not received much attention. To the best of the authors' knowledge, a thorough numerical analysis of the generic problem comparable with present study has not yet been attempted.

The main purpose of the paper is the validation of different modelling strategies for turbulent swirling flow of an incompressible fluid in an idealized swirl combustor.

Experiments have been performed and computations carried out for a water test rig, for a Reynolds number of 4,600 based on combustor inlet mean axial velocity and diameter. Two cases have been investigated, one low swirl and the other high swirl intensity. Measurements of time‐averaged velocity components and corresponding rms turbulence intensities were measured using laser Doppler anemometer, along radial traverses at different axial locations. In the three‐dimensional, unsteady computations, large eddy simulation (LES) and URANS (Unsteady Reynolds Averaged Navier‐Stokes Equations or Reynolds Averaged Numerical Simulations) RSMs (Reynolds‐stress models) are basically employed as modelling strategies for turbulence. To model subgrid‐scale turbulence for LES, the models due to Smagorinsky and Voke are used. No‐model LES and coarse‐grid direct numerical simulation computations are also performed for one of the cases.

The predictions are compared with the measurements and reveal that LES provided the best overall accuracy for all of the cases, whereas no significant difference between the Smagorinsky and Voke models are observed for the time‐averaged velocity components.

This paper provides additional valuable information on the performance of various modelling strategies for turbulent swirling flows.

The finite element solution of turbulent flows using a (k‐ε) turbulence model usually presents severe numerical difficulties. Develops an iterative (k‐L)‐predictor/ε‐corrector algorithm for overcoming this and solving the (k‐ε) turbulent models. The iterative scheme achieves convergence in L (length scale) which is proportional to (k1.5/ε). Numerical results indicate that the developed iterative algorithm is very robust.

To‐date, several segregated finite element algorithms have been proposed that solve the Navier—Stokes equations. These have considered only steady‐state cases. This paper describes the addition of the time‐dependent terms to one such segregated solution scheme. Several laminar flow examples have been computed and comparisons made to predictions obtained with both finite difference and finite volume solution schemes. The finite element results compare very well with the results from the other schemes, both in terms of accuracy and the qualitative behaviour of the iterative schemes.

Dynamic separator is an equipment having a rotor and static vanes and is used to separate solids from gas-solids flow based on size. Particle separation in a dynamic separator happens due to complex interchanges between multiple forces exerted in the separation zone. Currently, there is only limited knowledge concerning the working principles of separation. This paper aims to systematically study a dynamic separator using numerical models to get insights into particle separation.

The Lagrangian–Eulerian formulation is used to simulate gas-solid flow. Multiple frames of reference using stage interpolation are used to account for rotation. Periodic symmetry in the equipment is exploited to create a simplified numerical model. The predictions from the numerical model are compared against available experimental data.

The numerical results indicate that only when particle collision is included, the separation efficiency trend from the experiment is matched by numerical predictions. Further, it is shown that at the same range of rotor speeds where numerical results predict increased separation efficiency, the solid pressure due to particle collision also reaches its maximum value. The gas flow and particle behavior in the separator are explained in detail.

The importance of particle collision in separation is interesting because traditionally, particle separation is assumed to be influenced by three forces, namely, centrifugal force, drag force and gravity. The numerical results, however, point to the contribution by particle collision, in addition to the above three forces.

A finite element based numerical model is employed to obtain isothermal and heat transfer predictions for the case of turbulent flow with a decaying swirl component in a stationary circular pipe. An assessment is made on the quality of predictions based on the choice of turbulence modelling technique adopted to close the governing equations. In the present work the one‐equation, two‐equation and algebraic Reynolds stress turbulence models are employed. For the confined flow problem investigated, accurate prediction of the near‐wall conditions is essential. This is particularly the case for confined swirling flow where the variation of variables near the wall is often somewhat greater than encountered in pure axial flow. A finite element based near‐wall model is employed as an alternative to conventional techniques such as the use of the standard logarithmic functions. Of significance is the fact that flow predictions based on the use of the unidimensional finite element techniques are closer to experiment compared to the wall function based solutions for a given turbulence model. As expected, improvements in the flow predictions directly contribute to improved simulation of the thermal aspects of the problem.

Aging of the oil wells leads to a decrease in reservoir pressure and also to an increase in the water, gas and abrasive particles content. Therefore, there is a need for the oil pumps exploitation characteristics improvements. This paper aims to generate a valuable numerical model which will provide a useful tool to study various cases.

Computational fluid dynamics (CFD) analysis of the generation of so-called coherent structures of eddies and turbulence in the peripheral area of the vortex rotor mounted at the back side of centrifugal rotor was undertaken. After detailed analysis of the influence of the used turbulence models on the results, a hybrid turbulent model Detached Eddies Simulation (DES) was chosen as the most suitable.

Numerical control volume method with unsteady solver and DES turbulence model was proven to be valuable tool for flow analysis in the centrifugal pumps. Having in mind that DES turbulence model consumes much less computational time than large eddies turbulence model, this is a very useful fact that resulted from this research.

The proven numerical model is robust and reliable enough to become a standard method in simulating flow and other physical phenomena occurring in centrifugal pumps and similar turbo machines. This makes it possible to easily research different factors that influence their performances.

Comprehensive experimental and CFD study was performed which made it possible to conduct detailed validation and verification of described CFD model.

Refers to Montgomery and Fleeter (1996) who employed the finite‐analytic method of Chen et al. (1980) to study steady, two‐dimensional, inviscid, compressible, subsonic flow in a nozzle. Shows that, contrary to the statement made by Montgomery and Fleeter, their boundary conditions at the computational cell’s boundaries are not constructed from the particular solution to their equation (10). Deduces from a simple non‐linear second‐order ordinary differential equation that the finite or locally analytic method of Chen et al. (1980) only yields continuous but not differentiable solutions. Suggests a finite‐analytic method which provides continuous and differentiable solutions.