Search results1 – 2 of 2
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…
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…
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.