Abdel Dehbi - Academia.edu (original) (raw)
Papers by Abdel Dehbi
Proceedings of the 6th World Congress on Momentum, Heat and Mass Transfer, 2021
In this paper, we investigate Wall-Modeled Large Eddy Simulation (WMLES) in a stairmand high-effi... more In this paper, we investigate Wall-Modeled Large Eddy Simulation (WMLES) in a stairmand high-efficiency cyclone separator at two Reynolds numbers i.e. Re = 33,045-280,000. We compute the gas flow using the elliptic relaxation hybrid RANS/LES (ER-HRL). The model employs a wall integration linear eddy viscosity RANS model for the wall-adjacent region, and switches to the LES dynamic model in the bulk flow. At the lower Reynolds number i.e. Re = 33,045, we investigate the effect of varying cone bottom openingrendering three different cyclone configurationson cyclone performance. Flow statistics are reported at several locations across the cyclone axis where both mean and RMS values are observed. For high Reynolds number i.e. Re = 280,000, results are compared against LES and experimental databases from literature. Model predictions of mean flow are in good agreement with reference data, while higher-order moments i.e. RMS values are not very well predicted by the model despite following the same trend of experimental data. Results are in a global good agreement with LES and experimental data at a fraction of well-resolved LES CPU cost. This analysis will serve as a good basis for further investigation of cyclone grade efficiency using Lagrangian particle tracking.
Flow, Turbulence and Combustion, 2012
Agglomerate aerosols in a turbulent flow may be subjected to very high turbulent shear rates whic... more Agglomerate aerosols in a turbulent flow may be subjected to very high turbulent shear rates which through the generation of lift and drag can overcome the adhesive forces binding the constituents of an agglomerate together and cause it to break-up. This paper presents an analysis of the experimental measurements of the breakup of agglomerates between 0.1-10 μm in size, in a turbulent pipe flow followed by an expansion zone with a Reynolds numbers in the range 10 5 to 10 7. The analysis shows that even in wall bounded turbulence, the high turbulent shear stresses associated with the small scales of turbulence in the core can be the main source of breakup preceding any break-up that may occur by impaction at the wall. More importantly from these results, a computationally fast and efficient solution is obtained for the General Dynamic Equation (GDE) for agglomerate transport and breakup in highly turbulent flow. Furthermore the solution for the evolution of the aerosol size distribution is consistent with the experimental results. In the turbulent pipe flow section, the agglomerates are exposed continuously to turbulent shear stresses and experience more longer term breakup than in the expansion zone (following the pipe flow) where the exposure time is much less and break-up occurs instantaneously under the action of very high local turbulent shear stresses. The validity of certain approximations made in the model is considered. In particular, the inertia of the agglomerates characterised by a Stokes Number from 0.001 for the smallest particles up to 10 for 10 μm particles and the fluctuations of the turbulent
12th International Conference on Nuclear Engineering, Volume 2, 2004
ABSTRACT
Nuclear Engineering and Design - NUCL ENG DES, 2008
In Lagrangian particle dispersion modeling, the assumption that turbulence is isotropic everywher... more In Lagrangian particle dispersion modeling, the assumption that turbulence is isotropic everywhere yields erroneous predictions of particle deposition rates on walls, even in simple geometries. In this investigation, the stochastic particle tracking model in Fluent 6.2 is modified to include a better treatment of particle–turbulence interactions close to walls where anisotropic effects are significant. The fluid rms velocities in the boundary layer are computed using fits of DNS data obtained in channel flow. The new model is tested against correlations for particle removal rates in turbulent pipe flow and 90° bends. Comparison with experimental data is much better than with the default model. The model is also assessed against data of particle removal in the human mouth–throat geometry where the flow is decidedly three-dimensional. Here, the agreement with the data is reasonable, especially in view of the fact that the DNS fits used are those of channel flows, for lack of better al...
Participacion Del Ciemat En La 29 Reunion Anual De La Sociedad Nuclear Espanola Zaragoza 1 2 Y 3 De Octubre De 2003 2003 Pag 46, 2003
Recent Direct Numerical Simulations (DNS) (Marchioli et al.) and experimental data (Wang et al.) ... more Recent Direct Numerical Simulations (DNS) (Marchioli et al.) and experimental data (Wang et al.) have shown that inertial particles exhibit concentration peaks in isothermal turbulent boundary layers. It is therefore expected that turbulence will significantly enhance thermophoretic deposition in flows where walls are colder than the carrier gas. To correctly capture turbulent particle dispersion with active thermophoresis, a Lagrangian continuous random walk (CRW) model is developed. The particle tracking model uses 3D mean flow data obtained from the Fluent CFD code, as well as Eulerian statistics of instantaneous quantities computed from DNS databases. The turbulent fluid velocities at the current time step are related to those of the previous time step through a Markov chain based on the normalized Langevin equation (Iliopoulous & Hanratty 1999) which takes into account turbulence inhomogeneities. The model includes a correction to reduce the "spurious drift" of tracer-like particles which manifests itself in isothermal flows as an unphysical preferential concentration of fluid-like particles near the walls. This correction involves the addition of a "drift velocity" (Bocksell & Loth 2006) and yields improved results in isothermal flows such that tracerlike particles retain approximately uniform concentrations if introduced uniformly in the domain, and the deposition velocity of tracer particles is vanishingly small, as it should be.
The accurate prediction of particle transport is a primary safety issue. Tracking particles in La... more The accurate prediction of particle transport is a primary safety issue. Tracking particles in Lagrangian fashion can naturally be performed with CFD tools which provide the right framework to follow the paths of particles in complex geometries. The presence of turbulent structures in the fluid complicates the particle tracking problem considerably, because particle trajectories are no longer deterministic and additional modeling of the velocity fluctuations is needed. In the present investigation, a Lagrangian continuous random walk (CRW) model is developed to predict turbulent particle dispersion in wall-bounded flows with prevailing inhomogeneous turbulence. The particle model uses 3D mean flow data from the Fluent CFD code, as well as Eulerian statistics from Direct Numerical Simulation (DNS) databases. The turbulent fluid velocities are based on the non-dimensional Langevin equation. The model predictions are compared to the DNS data by Marchioli et al. (2007) who produced detailed statistics of velocity and transfer rates for classes of particles having Stokes numbers between 0.2 and 125 and dispersed in a parallel channel flow with Re τ =150. The model is in very good agreement with the DNS data for the various measures of particle dispersion. The predicted deposition rates are also in good agreement with the widely used experimental correlation of McCoy and Hanratty (1977) and Liu and Agarwal (1974).
In Lagrangian particle dispersion modeling, the assumption that turbulence is isotropic everywher... more In Lagrangian particle dispersion modeling, the assumption that turbulence is isotropic everywhere yields erroneous predictions of particle deposition rates on walls, even in simple geometries. In this investigation, the stochastic particle tracking model in FLUENT 6.2 is modified to include a better treatment of particle-turbulence interactions close to walls where anisotropic effects are significant. The fluid rms velocities in the boundary layer are computed using fits of DNS data obtained in channel flow. The new model is tested against correlations for particle removal rates in turbulent pipe flow and 90 o bends. Comparison with experimental data is much better than with the default model. The model is also assessed against data of particle removal in the human mouth-throat geometry where the flow is decidedly three dimensional. Here, the agreement with the data is reasonable, especially in view of the fact that the DNS fits used are those of channel flows, for lack of better alternatives. The CFD Best Practice Guidelines are followed to a large extent, in particular by using multiple grid resolutions and at least second order discretization schemes.
Proceedings of the 6th World Congress on Momentum, Heat and Mass Transfer, 2021
In this paper, we investigate Wall-Modeled Large Eddy Simulation (WMLES) in a stairmand high-effi... more In this paper, we investigate Wall-Modeled Large Eddy Simulation (WMLES) in a stairmand high-efficiency cyclone separator at two Reynolds numbers i.e. Re = 33,045-280,000. We compute the gas flow using the elliptic relaxation hybrid RANS/LES (ER-HRL). The model employs a wall integration linear eddy viscosity RANS model for the wall-adjacent region, and switches to the LES dynamic model in the bulk flow. At the lower Reynolds number i.e. Re = 33,045, we investigate the effect of varying cone bottom openingrendering three different cyclone configurationson cyclone performance. Flow statistics are reported at several locations across the cyclone axis where both mean and RMS values are observed. For high Reynolds number i.e. Re = 280,000, results are compared against LES and experimental databases from literature. Model predictions of mean flow are in good agreement with reference data, while higher-order moments i.e. RMS values are not very well predicted by the model despite following the same trend of experimental data. Results are in a global good agreement with LES and experimental data at a fraction of well-resolved LES CPU cost. This analysis will serve as a good basis for further investigation of cyclone grade efficiency using Lagrangian particle tracking.
Flow, Turbulence and Combustion, 2012
Agglomerate aerosols in a turbulent flow may be subjected to very high turbulent shear rates whic... more Agglomerate aerosols in a turbulent flow may be subjected to very high turbulent shear rates which through the generation of lift and drag can overcome the adhesive forces binding the constituents of an agglomerate together and cause it to break-up. This paper presents an analysis of the experimental measurements of the breakup of agglomerates between 0.1-10 μm in size, in a turbulent pipe flow followed by an expansion zone with a Reynolds numbers in the range 10 5 to 10 7. The analysis shows that even in wall bounded turbulence, the high turbulent shear stresses associated with the small scales of turbulence in the core can be the main source of breakup preceding any break-up that may occur by impaction at the wall. More importantly from these results, a computationally fast and efficient solution is obtained for the General Dynamic Equation (GDE) for agglomerate transport and breakup in highly turbulent flow. Furthermore the solution for the evolution of the aerosol size distribution is consistent with the experimental results. In the turbulent pipe flow section, the agglomerates are exposed continuously to turbulent shear stresses and experience more longer term breakup than in the expansion zone (following the pipe flow) where the exposure time is much less and break-up occurs instantaneously under the action of very high local turbulent shear stresses. The validity of certain approximations made in the model is considered. In particular, the inertia of the agglomerates characterised by a Stokes Number from 0.001 for the smallest particles up to 10 for 10 μm particles and the fluctuations of the turbulent
12th International Conference on Nuclear Engineering, Volume 2, 2004
ABSTRACT
Nuclear Engineering and Design - NUCL ENG DES, 2008
In Lagrangian particle dispersion modeling, the assumption that turbulence is isotropic everywher... more In Lagrangian particle dispersion modeling, the assumption that turbulence is isotropic everywhere yields erroneous predictions of particle deposition rates on walls, even in simple geometries. In this investigation, the stochastic particle tracking model in Fluent 6.2 is modified to include a better treatment of particle–turbulence interactions close to walls where anisotropic effects are significant. The fluid rms velocities in the boundary layer are computed using fits of DNS data obtained in channel flow. The new model is tested against correlations for particle removal rates in turbulent pipe flow and 90° bends. Comparison with experimental data is much better than with the default model. The model is also assessed against data of particle removal in the human mouth–throat geometry where the flow is decidedly three-dimensional. Here, the agreement with the data is reasonable, especially in view of the fact that the DNS fits used are those of channel flows, for lack of better al...
Participacion Del Ciemat En La 29 Reunion Anual De La Sociedad Nuclear Espanola Zaragoza 1 2 Y 3 De Octubre De 2003 2003 Pag 46, 2003
Recent Direct Numerical Simulations (DNS) (Marchioli et al.) and experimental data (Wang et al.) ... more Recent Direct Numerical Simulations (DNS) (Marchioli et al.) and experimental data (Wang et al.) have shown that inertial particles exhibit concentration peaks in isothermal turbulent boundary layers. It is therefore expected that turbulence will significantly enhance thermophoretic deposition in flows where walls are colder than the carrier gas. To correctly capture turbulent particle dispersion with active thermophoresis, a Lagrangian continuous random walk (CRW) model is developed. The particle tracking model uses 3D mean flow data obtained from the Fluent CFD code, as well as Eulerian statistics of instantaneous quantities computed from DNS databases. The turbulent fluid velocities at the current time step are related to those of the previous time step through a Markov chain based on the normalized Langevin equation (Iliopoulous & Hanratty 1999) which takes into account turbulence inhomogeneities. The model includes a correction to reduce the "spurious drift" of tracer-like particles which manifests itself in isothermal flows as an unphysical preferential concentration of fluid-like particles near the walls. This correction involves the addition of a "drift velocity" (Bocksell & Loth 2006) and yields improved results in isothermal flows such that tracerlike particles retain approximately uniform concentrations if introduced uniformly in the domain, and the deposition velocity of tracer particles is vanishingly small, as it should be.
The accurate prediction of particle transport is a primary safety issue. Tracking particles in La... more The accurate prediction of particle transport is a primary safety issue. Tracking particles in Lagrangian fashion can naturally be performed with CFD tools which provide the right framework to follow the paths of particles in complex geometries. The presence of turbulent structures in the fluid complicates the particle tracking problem considerably, because particle trajectories are no longer deterministic and additional modeling of the velocity fluctuations is needed. In the present investigation, a Lagrangian continuous random walk (CRW) model is developed to predict turbulent particle dispersion in wall-bounded flows with prevailing inhomogeneous turbulence. The particle model uses 3D mean flow data from the Fluent CFD code, as well as Eulerian statistics from Direct Numerical Simulation (DNS) databases. The turbulent fluid velocities are based on the non-dimensional Langevin equation. The model predictions are compared to the DNS data by Marchioli et al. (2007) who produced detailed statistics of velocity and transfer rates for classes of particles having Stokes numbers between 0.2 and 125 and dispersed in a parallel channel flow with Re τ =150. The model is in very good agreement with the DNS data for the various measures of particle dispersion. The predicted deposition rates are also in good agreement with the widely used experimental correlation of McCoy and Hanratty (1977) and Liu and Agarwal (1974).
In Lagrangian particle dispersion modeling, the assumption that turbulence is isotropic everywher... more In Lagrangian particle dispersion modeling, the assumption that turbulence is isotropic everywhere yields erroneous predictions of particle deposition rates on walls, even in simple geometries. In this investigation, the stochastic particle tracking model in FLUENT 6.2 is modified to include a better treatment of particle-turbulence interactions close to walls where anisotropic effects are significant. The fluid rms velocities in the boundary layer are computed using fits of DNS data obtained in channel flow. The new model is tested against correlations for particle removal rates in turbulent pipe flow and 90 o bends. Comparison with experimental data is much better than with the default model. The model is also assessed against data of particle removal in the human mouth-throat geometry where the flow is decidedly three dimensional. Here, the agreement with the data is reasonable, especially in view of the fact that the DNS fits used are those of channel flows, for lack of better alternatives. The CFD Best Practice Guidelines are followed to a large extent, in particular by using multiple grid resolutions and at least second order discretization schemes.