In particle systems,understanding mechanical interactions between particles is fundamental for accurately predicting material behavior under various conditions,especially when cohesive forces play a critical role.The ...In particle systems,understanding mechanical interactions between particles is fundamental for accurately predicting material behavior under various conditions,especially when cohesive forces play a critical role.The Discrete Element Method(DEM)is able to model these interactions based on material as well as contact properties.This paper evaluates the accuracy of DEM in predicting material behavior,for which a cost-effective cohesion model is implemented within an in-house DEM code.Simulations of wooden particles in a shear cell,with and without grease coating,are able to replicate the corresponding experimental results.The addition of the cohesion model results in a negligible increase in computational effort,which is of key importance for its application to larger systems.展开更多
Intraparticle models of thermal processes are essential in the Discrete Element Method (DEM) when particles are thermally thick. Accurately solving the intraparticle conservation equations using the finite volume meth...Intraparticle models of thermal processes are essential in the Discrete Element Method (DEM) when particles are thermally thick. Accurately solving the intraparticle conservation equations using the finite volume method requires sufficient spatial and temporal resolution, increasing simulation cost. This short communication proposes to replace the finite volume method (FVM) by a novel tabulated model that reduces the computational effort without compromising accuracy. The look-up database is generated from a set of single-particle simulations under different boundary conditions using the FVM. To assess the capability of the tabulated model to reproduce the results of the finite volume method, another single particle case with periodic functions for the fluid temperature is simulated. Results showed that tabulation provides accurate predictions while reducing computational cost by a factor of 3–10.展开更多
Discrete Ordinates Method(DOM)is a model for thermal radiation exchange in opaque media.In this study,the DOM formulation is employed within the framework of the Discrete Element Method coupled with Computational Flui...Discrete Ordinates Method(DOM)is a model for thermal radiation exchange in opaque media.In this study,the DOM formulation is employed within the framework of the Discrete Element Method coupled with Computational Fluid Dynamics(DEM-CFD),thus including full radiative heat exchange among the phases involved.This is done by adjusting the absorption coefficient,emission coefficient,and net radiative heat flux of particles by incorporating local porosity into equations.A key objective is to represent radiation propagation for different packing densities in packed beds accurately.The model is validated by comparing the results with available data from the literature for simulations with a P1 radiation model and corresponding experiments.The validation configuration is a heated box filled with spherical particles under vacuum conditions.As an application example,the radiative heat exchange between an enclosure at high temperature and moving layers of spherical particles concurrently passed by a gas in crossflow is studied.Three packing densities(dilute,moderate,and dense)are examined to evaluate radiation penetration into the particle ensemble.Convective and contact heat transfer are also considered.The DEM-CFD coupling is a non-resolved approach,where the influence of particles on the flow field is accounted for by momentum and energy source terms together with a porosity field(Averaged Volume Method(AVM)).Effect of convective,conductive and radiative heat transfer is analysed based on the evolution of incident radiation flux,spatial distributions of particle surface and fluid temperatures,and particle temperature histograms.It becomes obvious that radiation dominates the system,and that packing density defines the penetration depth of radiation.Conduction mainly leads to a smoothening of particle temperature distribution in the system.展开更多
Since many industrial applications rely on the processing of densely packed and moving granular ma-terial,obtaining bulk internal information on the particle movement inside the reactors is of great importance.Such in...Since many industrial applications rely on the processing of densely packed and moving granular ma-terial,obtaining bulk internal information on the particle movement inside the reactors is of great importance.Such information can be delivered by Positron Emission Particle Tracking(PEPT).By marking pellets with a positron-emitting radioisotope,the position of these tracer particles can be determined via the time-of-flight differences of the emitted gamma-ray pairs.The current paper proposes a PET-like detector system based on cost-effective organic plastic scintillators instead of the more common but expensive inorganic scintillators.This system is currently under construction and was tested for its resolution and efficiency in this simulation study.Using Monte Carlo simulations and the software toolkit Geant4,three different geometries(an empty glass box,a generic grate system,and a cubic box of 1 m3 completely filled with pellets)were investigated,leading to a spatial resolution in the millimeter range and an efficiency,defined as the ratio of reconstructed decay locations to simulated decays,of 2.7%,1.4%,and 0.3%.展开更多
Nowadays,the design of fixed packed bed reactors still relies on empirical correlations,which,especially for small tube to particle diameter ratios,are mostly too inaccurate because of the presence of wall effects.The...Nowadays,the design of fixed packed bed reactors still relies on empirical correlations,which,especially for small tube to particle diameter ratios,are mostly too inaccurate because of the presence of wall effects.Therefore,the simulation of fixed packed bed reactors plays an important role to predict and control the flow and process parameters in such,nowadays and in the future.Because of its straightforward applicability to non-uniform packings with particles of arbitrary shapes,the immersed boundary method(IBM)has advantages over other numerical methods and is used more and more frequently.This paper compares two approaches of IBMs for the simulation of fixed bed reactors with spherical shaped particles.The classic,smooth approach is compared to the straightforward to implement blocked-off method for velocity fields above the fixed bed for particle Reynolds numbers of 300 and 500.Results from experimental inline PIV-measurements of the reactor to be simulated serve as a basis for comparison.Very good agreement with the experiment is found for both simulation methodologies with higher resolutions,considering the more stable flow at a particle Reynolds number of 300.Differences in the different IBM approaches occurred for the more unsteady flow at a particle Reynolds number of 500.Compared to the blocked-off method,the smooth IBM reflects the formation of additional jets and recirculation zones better right above the bed,though increasing the fluid mesh resolution improves the accuracy of the blocked-off method.Overall,a more diffusive behaviour is found for the blocked-off simulations due to the stairstep representation,which is avoided by using interpolation stencils as in the smooth IBM.With higher mesh refinement in the blocked-off IBM this effect can be reduced,but this also increases the computational effort.展开更多
In industry,multiple hearth furnaces are used for the thermal treatment of particulate material.The current contribution concentrates on the experimental analysis of particle mechanics for a batch-operated single floo...In industry,multiple hearth furnaces are used for the thermal treatment of particulate material.The current contribution concentrates on the experimental analysis of particle mechanics for a batch-operated single floor of a multiple hearth furnace.The particles are agitated on the circular floor by a single,rotating rabble arm equipped with three flat rabble blades of 10 mm thickness.The blade angle,defined as the angle,which the blade is inclined against the tangential direction,is varied from 0°to 90°.A single layer of spherical polyoxymethylene(POM)particles with three different diameters(5,10 and 20 mm)is placed on the floor.To analyze the results,two parameters have been extracted from image analysis when the bed of particles is agitated,first,the area not covered by particles and second,the frequency distribution of the mean distance among the particles.The particle free surface area increases with the inclination of the blades.The evolution of the particle free surface area differs for the different particle diameters.In general,the maximum particle free area for all blade angles is the largest for the 5 mm particles followed by the 20 mm particles.For the 10 mm particles,the particle free surface area starts for a blade angle of 0°at larger values than for the 20 mm particles but the values fall below the values for the 20 mm particles for larger blade angles.The reason for this behavior is discussed in detail.The mean distance among the particles is a parameter characterizing the length scales dominating the effects on the floor.The frequency distribution of the mean distance among particles provides infor-mation about the morphology of the particle bulk,for example,whether the free surface area is inter-spersed with particles.展开更多
The simulation of industry-scale reactive bulks is challenging due to the complex interaction between fluid and particles.The particles in the bulk and their interaction with the fluid flow can be described by combine...The simulation of industry-scale reactive bulks is challenging due to the complex interaction between fluid and particles.The particles in the bulk and their interaction with the fluid flow can be described by combined Discrete Element Method-Computational Fluid Dynamics(DEM-CFD)models.However,the computational cost of the Finite Volume(FV)methods deployed in these models can become prohibi-tively expensive,especially for high inner-particle resolution.Single particle Reduced Models(RMs)can be used to achieve both fast and accurate descriptions of the processes in each particle.As an example of bulk systems comprising heat and mass transfer,we compared FV and RM simulations for the drying of wood chips in a bulk reactor.A manifold-based nonlinear interpolation was applied to resolve changing boundary conditions for the RM.Our simulations showed that RMs provide accurate values for the thermodynamic state variables of the particles.Furthermore,the time required for the bulk simulation was reduced by 67%with the RMs.It is evident that simulations with high inner-particle resolution can be accelerated by RMs if manifold-based nonlinear interpolation is used to address changing boundary conditions.展开更多
This work is the second part of a simulation study investigating the processing of densely packed and moving granular assemblies by positron emission particle tracking (PEPT).Since medical positron emission tomography...This work is the second part of a simulation study investigating the processing of densely packed and moving granular assemblies by positron emission particle tracking (PEPT).Since medical positron emission tomography (PET) scanners commonly used for PEPT are very expensive,a PET-like detector system based on cost-effective organic plastic scintillator bars is being developed and tested for its capabilities.In this context,the spatial resolution of a resting positron source,a source moving on a freely designed model path,and a particle motion given by a discrete element method (DEM) simulation is studied using Monte Carlo simulations and the software toolkit Geant4.This not only extended the simulation and reconstruction to a moving source but also significantly improved the spatial resolution compared to previous work by adding oversampling and iteration to the reconstruction algorithm.Furthermore,in the case of a source following a trajectory developed from DEM simulations,a very good resolution of about 1 mm in all three directions and an average 3D deviation between simulated and reconstructed events of 2.3 mm could be determined.Thus,the resolution for realistic particle motion within the generic grate system (which is the test rig for further experimental studies) is well below the smallest particle size.The simulation of the dependence of the reconstruction accuracy on tracer particle location revealed a nearly constant efficiency within the entire detector system,which demonstrates that boundary effects can be neglected.展开更多
Taking into account the complex shape of particles in discrete element method (DEM) simulations of large-scale granular systems is computationally demanding due to the time-consuming contact detection algorithms for p...Taking into account the complex shape of particles in discrete element method (DEM) simulations of large-scale granular systems is computationally demanding due to the time-consuming contact detection algorithms for polyhedral particles. In this short communication, a novel approach that locally resolves the particle shapes where needed and uses a simplified representation elsewhere, to accelerate simulations without compromising accuracy, is presented. For this purpose, a method employing a smooth transition of the particle shape representation from analytical spheres to shape-resolving polyhedra is introduced in DEM. The feasibility and correct implementation of this approach are demonstrated through simulations of hopper discharge involving spherical and dodecahedral particles from a flat bottom silo or shaft kiln. The model capabilities, in terms of accuracy as well as reduction in computational effort, are quantified for a moving bed with continuous outflow.展开更多
We treat the accurate simulation of the calcination reaction in particles,where the particles are large and,thus,the inner-particle processes must be resolved.Because these processes need to be described with coupled ...We treat the accurate simulation of the calcination reaction in particles,where the particles are large and,thus,the inner-particle processes must be resolved.Because these processes need to be described with coupled partial differential equations(PDEs)that must be solved numerically,the computation times for a single particle are too high for use in simulations that involve many particles.Simulations of this type arise when the Discrete Element Method(DEM)is combined with Computational Fluid Dynamics(CFD)to investigate industrial systems such as quicklime production in lime shaft kilns.We show that,based on proper orthogonal decomposition and Galerkin projection,reduced models can be derived for single particles that provide the same spatial and temporal resolution as the original PDE models at a considerably reduced computational cost.Replacing the finite volume particle models with the reduced models results in an overall reduction of the reactor simulation time by about 40%for the sample system treated here.展开更多
Discrete Element Method-Computational Fluid Dynamics(DEM/CFD)simulations of industrial-scale granular systems employ spatial averaging(porous media approach)for the fluid-particle interaction in the whole domain,which...Discrete Element Method-Computational Fluid Dynamics(DEM/CFD)simulations of industrial-scale granular systems employ spatial averaging(porous media approach)for the fluid-particle interaction in the whole domain,which can lead to poor accuracy,for instance at flow inlets,as local particle bulk morphology is not resolved.This paper presents an approach where the interstitial flow in crucial areas with large gradients can be resolved locally in an otherwise unresolved domain,so that a mixed resolved-unresolved method is realized.As a generic example to show the feasibility and performance of the new approach,the inflow of ambient air into a flat-bottom hopper through a narrow orifice is investigated.In an experimental setup,the vertical profile of the pressure decay through the inlet and across the packing is chosen for com-parison with respective simulations.Results obtained with the conventional porous media method and the locally resolved approach are compared to these experiments for varying volume flow rates and for two different particle shapes.Spheres of different size as well as dodecahedrons are examined.It is found that although averaging methods already provide good approximations,the locally resolved method can improve the result especially when conventional drag laws are not applicable due to wall effects or if large velocity gradients exist.展开更多
基金funded by the European Commission under the“Horizon Europe(Grant agreement 101058429)”project,specifically within the“Maximise H2 Enrichment in Direct Reduction Shaft Furnaces”(MaxH2DR)project.
文摘In particle systems,understanding mechanical interactions between particles is fundamental for accurately predicting material behavior under various conditions,especially when cohesive forces play a critical role.The Discrete Element Method(DEM)is able to model these interactions based on material as well as contact properties.This paper evaluates the accuracy of DEM in predicting material behavior,for which a cost-effective cohesion model is implemented within an in-house DEM code.Simulations of wooden particles in a shear cell,with and without grease coating,are able to replicate the corresponding experimental results.The addition of the cohesion model results in a negligible increase in computational effort,which is of key importance for its application to larger systems.
基金funded by the Deutsche Forschungsgemeinschaft(DFG,German Research Foundation)–Project-ID 422037413–TRR 287.(Gefördert durch die Deutsche Forschungsgemeinschaft(DFG)–Projektnummer 422037413–TRR 287).
文摘Intraparticle models of thermal processes are essential in the Discrete Element Method (DEM) when particles are thermally thick. Accurately solving the intraparticle conservation equations using the finite volume method requires sufficient spatial and temporal resolution, increasing simulation cost. This short communication proposes to replace the finite volume method (FVM) by a novel tabulated model that reduces the computational effort without compromising accuracy. The look-up database is generated from a set of single-particle simulations under different boundary conditions using the FVM. To assess the capability of the tabulated model to reproduce the results of the finite volume method, another single particle case with periodic functions for the fluid temperature is simulated. Results showed that tabulation provides accurate predictions while reducing computational cost by a factor of 3–10.
基金funded by the Deutsche Forschungsgemeinschaft(DFG,German Research Foundation)–Project-ID 422037413–TRR 287Gefördert durch die Deutsche Forschungsgemeinschaft(DFG)–Projektnummer 422037413–TRR 287.
文摘Discrete Ordinates Method(DOM)is a model for thermal radiation exchange in opaque media.In this study,the DOM formulation is employed within the framework of the Discrete Element Method coupled with Computational Fluid Dynamics(DEM-CFD),thus including full radiative heat exchange among the phases involved.This is done by adjusting the absorption coefficient,emission coefficient,and net radiative heat flux of particles by incorporating local porosity into equations.A key objective is to represent radiation propagation for different packing densities in packed beds accurately.The model is validated by comparing the results with available data from the literature for simulations with a P1 radiation model and corresponding experiments.The validation configuration is a heated box filled with spherical particles under vacuum conditions.As an application example,the radiative heat exchange between an enclosure at high temperature and moving layers of spherical particles concurrently passed by a gas in crossflow is studied.Three packing densities(dilute,moderate,and dense)are examined to evaluate radiation penetration into the particle ensemble.Convective and contact heat transfer are also considered.The DEM-CFD coupling is a non-resolved approach,where the influence of particles on the flow field is accounted for by momentum and energy source terms together with a porosity field(Averaged Volume Method(AVM)).Effect of convective,conductive and radiative heat transfer is analysed based on the evolution of incident radiation flux,spatial distributions of particle surface and fluid temperatures,and particle temperature histograms.It becomes obvious that radiation dominates the system,and that packing density defines the penetration depth of radiation.Conduction mainly leads to a smoothening of particle temperature distribution in the system.
基金funded by the DeutscheForschungsgemeinschaft(DFG,German Research Foundation)through 422037413-CRC/TRR 287"BULK-REACTION"。
文摘Since many industrial applications rely on the processing of densely packed and moving granular ma-terial,obtaining bulk internal information on the particle movement inside the reactors is of great importance.Such information can be delivered by Positron Emission Particle Tracking(PEPT).By marking pellets with a positron-emitting radioisotope,the position of these tracer particles can be determined via the time-of-flight differences of the emitted gamma-ray pairs.The current paper proposes a PET-like detector system based on cost-effective organic plastic scintillators instead of the more common but expensive inorganic scintillators.This system is currently under construction and was tested for its resolution and efficiency in this simulation study.Using Monte Carlo simulations and the software toolkit Geant4,three different geometries(an empty glass box,a generic grate system,and a cubic box of 1 m3 completely filled with pellets)were investigated,leading to a spatial resolution in the millimeter range and an efficiency,defined as the ratio of reconstructed decay locations to simulated decays,of 2.7%,1.4%,and 0.3%.
基金funded by the Deutsche Forschungsgemeinschaft(DFG,German Research Foundation)-Project-ID 422037413-TRR 287Gefordert durch die Deutsche Forschungsgemeinschaft(DFG)-Projektnummer 422037413-TRR 287.
文摘Nowadays,the design of fixed packed bed reactors still relies on empirical correlations,which,especially for small tube to particle diameter ratios,are mostly too inaccurate because of the presence of wall effects.Therefore,the simulation of fixed packed bed reactors plays an important role to predict and control the flow and process parameters in such,nowadays and in the future.Because of its straightforward applicability to non-uniform packings with particles of arbitrary shapes,the immersed boundary method(IBM)has advantages over other numerical methods and is used more and more frequently.This paper compares two approaches of IBMs for the simulation of fixed bed reactors with spherical shaped particles.The classic,smooth approach is compared to the straightforward to implement blocked-off method for velocity fields above the fixed bed for particle Reynolds numbers of 300 and 500.Results from experimental inline PIV-measurements of the reactor to be simulated serve as a basis for comparison.Very good agreement with the experiment is found for both simulation methodologies with higher resolutions,considering the more stable flow at a particle Reynolds number of 300.Differences in the different IBM approaches occurred for the more unsteady flow at a particle Reynolds number of 500.Compared to the blocked-off method,the smooth IBM reflects the formation of additional jets and recirculation zones better right above the bed,though increasing the fluid mesh resolution improves the accuracy of the blocked-off method.Overall,a more diffusive behaviour is found for the blocked-off simulations due to the stairstep representation,which is avoided by using interpolation stencils as in the smooth IBM.With higher mesh refinement in the blocked-off IBM this effect can be reduced,but this also increases the computational effort.
基金funded by the Deutsche Forschungsgemeinschaft(DFG,German Research Foundation)-Project-ID 422037413-TRR 287.
文摘In industry,multiple hearth furnaces are used for the thermal treatment of particulate material.The current contribution concentrates on the experimental analysis of particle mechanics for a batch-operated single floor of a multiple hearth furnace.The particles are agitated on the circular floor by a single,rotating rabble arm equipped with three flat rabble blades of 10 mm thickness.The blade angle,defined as the angle,which the blade is inclined against the tangential direction,is varied from 0°to 90°.A single layer of spherical polyoxymethylene(POM)particles with three different diameters(5,10 and 20 mm)is placed on the floor.To analyze the results,two parameters have been extracted from image analysis when the bed of particles is agitated,first,the area not covered by particles and second,the frequency distribution of the mean distance among the particles.The particle free surface area increases with the inclination of the blades.The evolution of the particle free surface area differs for the different particle diameters.In general,the maximum particle free area for all blade angles is the largest for the 5 mm particles followed by the 20 mm particles.For the 10 mm particles,the particle free surface area starts for a blade angle of 0°at larger values than for the 20 mm particles but the values fall below the values for the 20 mm particles for larger blade angles.The reason for this behavior is discussed in detail.The mean distance among the particles is a parameter characterizing the length scales dominating the effects on the floor.The frequency distribution of the mean distance among particles provides infor-mation about the morphology of the particle bulk,for example,whether the free surface area is inter-spersed with particles.
基金funded by the Deutsche For-schungsgemeinschaft(DFG,German Research Foundation)-Project-ID 422037413-TRR 287.Gefordert durch die Deutsche Forschungsgemeinschaft(DFG)-Projektnummer 422037413-TRR 287.
文摘The simulation of industry-scale reactive bulks is challenging due to the complex interaction between fluid and particles.The particles in the bulk and their interaction with the fluid flow can be described by combined Discrete Element Method-Computational Fluid Dynamics(DEM-CFD)models.However,the computational cost of the Finite Volume(FV)methods deployed in these models can become prohibi-tively expensive,especially for high inner-particle resolution.Single particle Reduced Models(RMs)can be used to achieve both fast and accurate descriptions of the processes in each particle.As an example of bulk systems comprising heat and mass transfer,we compared FV and RM simulations for the drying of wood chips in a bulk reactor.A manifold-based nonlinear interpolation was applied to resolve changing boundary conditions for the RM.Our simulations showed that RMs provide accurate values for the thermodynamic state variables of the particles.Furthermore,the time required for the bulk simulation was reduced by 67%with the RMs.It is evident that simulations with high inner-particle resolution can be accelerated by RMs if manifold-based nonlinear interpolation is used to address changing boundary conditions.
基金funded by the Deutsche Forschungsgemeinschaft (DFG,German Research Foundation)-Project-ID 422037413-TRR 287.
文摘This work is the second part of a simulation study investigating the processing of densely packed and moving granular assemblies by positron emission particle tracking (PEPT).Since medical positron emission tomography (PET) scanners commonly used for PEPT are very expensive,a PET-like detector system based on cost-effective organic plastic scintillator bars is being developed and tested for its capabilities.In this context,the spatial resolution of a resting positron source,a source moving on a freely designed model path,and a particle motion given by a discrete element method (DEM) simulation is studied using Monte Carlo simulations and the software toolkit Geant4.This not only extended the simulation and reconstruction to a moving source but also significantly improved the spatial resolution compared to previous work by adding oversampling and iteration to the reconstruction algorithm.Furthermore,in the case of a source following a trajectory developed from DEM simulations,a very good resolution of about 1 mm in all three directions and an average 3D deviation between simulated and reconstructed events of 2.3 mm could be determined.Thus,the resolution for realistic particle motion within the generic grate system (which is the test rig for further experimental studies) is well below the smallest particle size.The simulation of the dependence of the reconstruction accuracy on tracer particle location revealed a nearly constant efficiency within the entire detector system,which demonstrates that boundary effects can be neglected.
基金funded by the Deutsche Forschungsgemeinschaft(DFG,German Research Foundation)-Project-ID 422037413-TRR 287Gefördert durch die Deutsche Forschungsgemeinschaft(DFG)-Projektnummer 422037413-TRR 287.
文摘Taking into account the complex shape of particles in discrete element method (DEM) simulations of large-scale granular systems is computationally demanding due to the time-consuming contact detection algorithms for polyhedral particles. In this short communication, a novel approach that locally resolves the particle shapes where needed and uses a simplified representation elsewhere, to accelerate simulations without compromising accuracy, is presented. For this purpose, a method employing a smooth transition of the particle shape representation from analytical spheres to shape-resolving polyhedra is introduced in DEM. The feasibility and correct implementation of this approach are demonstrated through simulations of hopper discharge involving spherical and dodecahedral particles from a flat bottom silo or shaft kiln. The model capabilities, in terms of accuracy as well as reduction in computational effort, are quantified for a moving bed with continuous outflow.
文摘We treat the accurate simulation of the calcination reaction in particles,where the particles are large and,thus,the inner-particle processes must be resolved.Because these processes need to be described with coupled partial differential equations(PDEs)that must be solved numerically,the computation times for a single particle are too high for use in simulations that involve many particles.Simulations of this type arise when the Discrete Element Method(DEM)is combined with Computational Fluid Dynamics(CFD)to investigate industrial systems such as quicklime production in lime shaft kilns.We show that,based on proper orthogonal decomposition and Galerkin projection,reduced models can be derived for single particles that provide the same spatial and temporal resolution as the original PDE models at a considerably reduced computational cost.Replacing the finite volume particle models with the reduced models results in an overall reduction of the reactor simulation time by about 40%for the sample system treated here.
基金funded by the Deutsche Forschungsgemeinschaft(DFG,German Research Foundation)-Project-ID 422037413-TRR 287.
文摘Discrete Element Method-Computational Fluid Dynamics(DEM/CFD)simulations of industrial-scale granular systems employ spatial averaging(porous media approach)for the fluid-particle interaction in the whole domain,which can lead to poor accuracy,for instance at flow inlets,as local particle bulk morphology is not resolved.This paper presents an approach where the interstitial flow in crucial areas with large gradients can be resolved locally in an otherwise unresolved domain,so that a mixed resolved-unresolved method is realized.As a generic example to show the feasibility and performance of the new approach,the inflow of ambient air into a flat-bottom hopper through a narrow orifice is investigated.In an experimental setup,the vertical profile of the pressure decay through the inlet and across the packing is chosen for com-parison with respective simulations.Results obtained with the conventional porous media method and the locally resolved approach are compared to these experiments for varying volume flow rates and for two different particle shapes.Spheres of different size as well as dodecahedrons are examined.It is found that although averaging methods already provide good approximations,the locally resolved method can improve the result especially when conventional drag laws are not applicable due to wall effects or if large velocity gradients exist.
