Simulation Of Particle-laden Flow Research Articles

Simulation of particle-laden flows and erosion in an axial fan stage considering the relative position of the blades

Abstract Axial fans are vital accessories in aircraft ventilation systems, but, they may experience erosion from particulate flows, causing a decline in effectiveness over time. This study investigated the trajectories of two types of sand particles and erosion in an axial fan stage, considering the relative position of the blades facing the inlet guide vanes. The movement of particles was simulated using an in-house code that implements a Lagrangian approach along with a stochastic particle-eddy interaction model. The flow field was solved separately and the flow data was transferred to the particle trajectory code. The finite element method allowed for the tracking of particles through the computational cells and accurate determination of their impact positions. A semi-empirical erosion correlation was used to evaluate the local erosion rates, mass removal, and geometry deterioration. As a result, the rotor exhibits a high frequency of impacts and significant erosion on the leading edge of the blade, extending to the upper corner of the pressure side and blade tip, as well as the front of the suction side. In the inlet guide vane, the erosion is spread out along the entire pressure side but at lower erosion rates compared to the rotor blade. The erosion patterns obtained at different pitch-wise positions were cumulated to get better representation of erosion patterns. After being exposed to MIL-E5007E sand (0–1000 $\unicode{x03BC}$ m) at the highest concentration for 10 hours, the blade experienced a reduction of a 0.29% in mass, a 0.45% decrease in tip chord, and a 0.23% increase in tip clearance. On the other hand, AC-coarse sand (0–200 μm) resulted in a 0.23% decrease in blade mass, a 0.4% reduction in tip chord, and a 0.16% increase in tip clearance. The data that is available can be used to monitor the lifespan of axial fans of similar design and select appropriate coatings to protect against erosion.

A flow-reconstruction based approach for the computation of hydrodynamic stresses on immersed body surface

This study focuses on obtaining reliable hydrodynamic stresses on the surface of rigid bodies in the simulation of fluid-solid interactions using the diffuse-interface immersed boundary method (IBM). In order to obtain the pressure and velocity gradient information on the immersed boundary, we employ a model to reconstruct the near-boundary fluid flow along the surface normal direction. Different from most existing flow reconstruction models in diffuse-interface IBMs based on the boundary layer approximation, we propose a new reconstruction model relying only on the non-slip velocity and local flow information. With this improvement, the present reconstruction model is theoretically applicable to fluid flows with intermediate Reynolds numbers (with the existence of a laminar boundary layer) and low Reynolds numbers (in the absence of a boundary layer). To examine the validity of the present reconstruction model, we conduct a series of simulations, including the uniform flow past a fixed body and the settling of a rigid body in a quiescent fluid. It is demonstrated that the present reconstruction model has a better performance than existing models in predicting hydrodynamic stresses on the immersed boundary within a wide range of Reynolds numbers from O(1) to O(100). The capability and robustness of the proposed model demonstrate its potential to address fluid-solid interaction problems with a-priori unknown or time-varying Reynolds numbers, such as fully resolved simulations of particle laden flows.

Turbulence Unsteadiness Drives Extreme Clustering.

We show that the unsteadiness of turbulence has a drastic effect on turbulence parameters and in particle cluster formation. To this end we use direct numerical simulations of particle laden flows with a steady forcing that generates an unsteady large-scale flow. Particle clustering correlates with the instantaneous Taylor-based flow Reynolds number, and anticorrelates with its instantaneous turbulent energy dissipation constant. A dimensional argument for these correlations is presented. In natural flows, unsteadiness can result in extreme particle clustering, which is stronger than the clustering expected from averaged inertial turbulence effects.

Simulation of Particle Trajectories in Gas Turbine Components and Assessment of Unsteady Effects Using an Efficient Eulerian-Lagrangian Technique

In recent years, CFD has proven to be a very useful asset to help with predicting complex flows in a wide range of situations, including multiphase and gas-particle flows. On this track, numerical modelling of particle-laden flows in multistage turbomachinery has become an important step in helping to analyse the behaviour of a discrete phase in gas turbines. Furthermore, unsteady effects due, for example, to rotor–stator interaction may have an effect on trajectories and capture efficiencies of the discrete phase. Unfortunately, computational times for transient simulations can be exceedingly high, especially if a discrete-phase needs also to be simulated. For this reason, this work reports a new method for the efficient and accurate simulation of particle-laden flows in gas turbine engines components. The Harmonic Balance Method is exploited to gain orders of magnitude speedup exploiting the idea that once the flow field has been embedded in the spectral basis, it can be reconstructed at any desired time. In this way, not only can the computational time needed to reach convergence of the flow field be dramatically reduced, but there is also no need to keep simulating the flow field during particle tracking. On the contrary, the continuous phase field can be retrieved at any desired time through flow reconstruction. This technique is conceptually simple, but, to the authors’ knowledge, has never been applied so far in particle-laden flow simulations and represents a novelty in the field. First, the implementation of the method is described, and details are given on how phase-lagged boundary conditions can be applied to flow and particles to further speed up the calculation. Then, some relevant case studies are presented to highlight the performance of the method.

A neural network based framework to model particle rebound and fracture

The current state of the art approach in the simulation of particle-laden flow in turbomachinery is to handle particle–wall interactions via rebound and erosion models. Rebound models often require a priori parameter tuning to match experimental measurements. Moreover, the actual stochastic nature of the rebound is neglected, and the particle is assumed not to fracture upon impact. However, this affects the resulting particle trajectories and is particularly critical at high (normal) impact velocities, where particles in typical aero engine flow exhibit a high probability of fracture, as illustrated in our previous work. In this work, we propose a method to develop a generalized rebound model which is parameter-free for the user and considers the stochasticity of the rebound. To this end, state of the art methods from function approximation, more precisely, deep dense neural networks are employed. The networks are trained through a supervised learning approach, where the neural network maps the impacting particles’ characteristics to its new particle trajectory after rebound. For this, we present an efficient method to predict probability distributions in a supervised learning context without a priori parameter tuning of known PDFs. In a second step, we extend the network to account for particle fracture, where the particle breakage is based on a fracture probability distribution to determine whether a particle breaks. The performance of the proposed framework is illustrated by the use of experimental measurements of the statistical rebound of sand particles in an erosion test rig specifically designed to match flow and impact conditions (including particle fracture) as well as particle sizes in aero engine compressors.

Lift-off of multiple particles in a narrow channel

In this work, we perform simulations of particle laden flow in a wide and long narrow channel in a Newtonian fluid. Simulations are performed for mono-sized and equal density spheres with varying Archimedes and Reynolds number. In the simulations, different phases of particle transport - rolling, saltation and suspension are observed. During simulations the average bed height is monitored and its steady state value is used for proposing a correlation between solids volume fraction (ϕ), shear Reynolds number (Res) and Archimedes number (Ar). This correlation is used to predict the critical shear Reynolds number for particle lift-off and a condition at which particles would occupy the whole channel at a given Archimedes number. The value of this critical Reynolds number is compared with the critical Reynolds number for single particle lift-off.

High-order overset grid method for detecting particle impaction on a cylinder in a cross flow

An overset grid method was developed to investigate the interaction between a particle-laden flow and a circular cylinder. The method is implemented in the Pencil Code, a high-order finite-difference code for compressible flow simulation. High-order summation-by-parts operators were used at the cylinder boundary, and both bi-linear Lagrangian and bi-quadratic spline interpolation were used to communicate between the Cartesian background grid and the body-conformal cylindrical grid. The performance of the overset grid method was assessed to benchmark cases of steady and unsteady flows past a cylinder. Results show high-order accuracy and good agreement to the literature. Particle-laden flow simulations were performed, with inertial point particles impacting on a cylinder. The simulations reproduced results from the literature at a significantly reduced cost. Further, an investigation into blockage effects on particle impaction revealing that the previously published DNS data is less accurate than assumed for particles with very small Stokes numbers.

Effects of Polarization on Particle-Laden Flows.

Simulations of particle-laden flow with dielectric particles are carried out with varying levels of electrical charging and particle polarization. Simulation results reveal three distinct flow regions. For low particle charge and polarizability, flow is nearly symmetric and nonmeandering. For strong charging and polarization, particles form a continuous and tightly clustered sheet close to one of the walls. Between these extremes, particles form localized particle-rich regions, around which the gas executes a meandering flow. These results indicate that polarization can lead to qualitative changes in the characteristics of particle-laden flows subject to tribocharging.

An improved ghost-cell sharp interface immersed boundary method with direct forcing for particle laden flows

In this paper, an accurate and stable sharp interface immersed boundary method(IBM) is presented for the direct numerical simulation of particle laden flows. The current IBM method is based on the direct-forcing method by incorporating the ghost-cell approach implicitly. An important feature of this IBM is the sharp representation of the solid surface, contrary to other variants of IBM for freely moving particles in which the solid surface is diffuse. Moreover, a correction of the diameter is not necessary for obtaining accurate results. The current ghost-cell IBM is stable because spurious oscillations incurred due to discontinuity in the pressure and velocity field in moving particle simulations is avoided. An algorithm for accurate torque computation is developed. The proposed algorithm is verified by comparison to an analytical expression and is shown to give a substantial improvement over the existing method. Finally, the present IBM is validated for various test cases of single and multi-particle systems and is shown to be accurate and robust for a wide range of flow conditions.

Direct numerical simulation of particle laden flow in a human airway bifurcation model

During the delivery of inhaled medicines, and depending on the size distribution of the particles in the formulation, airway bifurcations are areas of preferential deposition. Previous studies of laminar flow through airway bifurcations point to an interplay of inertial and centrifugal forces that leads to rich flow phenomena and controls particle deposition patterns. However, recent computational studies have shown that the airflow in the upper human airways is turbulent during much of the respiratory cycle. The question of how the presence of turbulence modifies these effects remains open. In this study, we perform for the first time Direct Numerical Simulations (DNS) of fully developed turbulent flow through a single human airway bifurcation model, emulating steady prolonged inspiration and expiration. We use the rich information obtained from the DNS in order to identify key structures in the flow field and scrutinize their role in determining deposition patterns in the bifurcation. We find that the vortical structures present in the bifurcation during expiration differ from those identified during inspiration. While Dean vortices are present in both cases, a set of three dimensional “carinal vortices” are identified only during expiration. A set of laminar simulations in the same geometries, but at lower Reynolds numbers, allow us to identify key differences in aerosol deposition patterns between laminar and turbulent respiration. We also report deposition fractions for representative Stokes numbers for both laminar and turbulent conditions. Given the suspected role of external mechanical stress on the airway epithelium in determining mucus clearance and chronic disease development, here we report wall shear stress distributions for both the turbulent and laminar cases. Finally, we also perform Large Eddy Simulations (LES) and Reynolds-Averaged Navier-Stokes (RANS) simulations for the same configuration in order to asses their performance as compared to DNS. We find that LES and RANS perform well and that they are able to capture the key characteristics of the flow field. The agreement between DNS and RANS holds true only for the mean flow field, which is primarily influenced by curvature effects.

A parallel fictitious domain method for the interface-resolved simulation of particle-laden flows and its application to the turbulent channel flow

ABSTRACTA parallel direct-forcing (DF) fictitious domain (FD) method for the simulation of particulate flows is reported in this paper. The parallel computing strategies for the solution of flow fields and particularly the distributed Lagrange multiplier are presented, and the high efficiency of the parallel code is demonstrated. The new code is then applied to study the effects of particle density (or particle inertia) on the turbulent channel flow. The results show that the large-scale vortices are weakened more severely, and the flow friction drag increases first and then reduces, as particle inertia is increased.

Dynamic coupling of fluid and structural mechanics for simulating particle motion and interaction in high speed compressible gas particle laden flow

In this work, structural finite element analyses of particles moving and interacting within high speed compressible flow are directly coupled to computational fluid dynamics and heat transfer analyses to provide more detailed and improved simulations of particle laden flow under these operating conditions. For a given solid material model, stresses and displacements throughout the solid body are determined with the particle–particle contact following an element to element local spring force model and local fluid induced forces directly calculated from the finite volume flow solution. Plasticity and particle deformation common in such a flow regime can be incorporated in a more rigorous manner than typical discrete element models where structural conditions are not directly modeled. Using the developed techniques, simulations of normal collisions between two 1mm radius particles with initial particle velocities of 50–150m/s are conducted with different levels of pressure driven gas flow moving normal to the initial particle motion for elastic and elastic–plastic with strain hardening based solid material models. In this manner, the relationships between the collision velocity, the material behavior models, and the fluid flow and the particle motion and deformation can be investigated. The elastic–plastic material behavior results in post collision velocities 16–50% of their pre-collision values while the elastic-based particle collisions nearly regained their initial velocity upon rebound. The elastic–plastic material models produce contact forces less than half of those for elastic collisions, longer contact times, and greater particle deformation. Fluid flow forces affect the particle motion even at high collision speeds regardless of the solid material behavior model. With the elastic models, the collision force varied little with the strength of the gas flow driver. For the elastic–plastic models, the larger particle deformation and the resulting increasingly asymmetric loading lead to growing differences in the collision force magnitudes and directions as the gas flow strength increased. The coupled finite volume flow and finite element structural analyses provide a capability to capture the interdependencies between the interaction of the particles, the particle deformation, the fluid flow and the particle motion.

Recent advances in the simulation of particle-laden flows

A substantial number of algorithms exists for the simulation of moving particles suspended in fluids. However, finding the best method to address a particular physical problem is often highly non-trivial and depends on the properties of the particles and the involved fluid(s) together. In this report, we provide a short overview on a number of existing simulation methods and provide two state of the art examples in more detail. In both cases, the particles are described using a Discrete Element Method (DEM). The DEM solver is usually coupled to a fluid-solver, which can be classified as grid-based or mesh-free (one example for each is given). Fluid solvers feature different resolutions relative to the particle size and separation. First, a multicomponent lattice Boltzmann algorithm (mesh-based and with rather fine resolution) is presented to study the behavior of particle stabilized fluid interfaces and second, a Smoothed Particle Hydrodynamics implementation (mesh-free, meso-scale resolution, similar to the particle size) is introduced to highlight a new player in the field, which is expected to be particularly suited for flows including free surfaces.

An improved immersed boundary method with direct forcing for the simulation of particle laden flows

An efficient approach for the simulation of finite-size particles with interface resolution was presented by Uhlmann [M. Uhlmann, An immersed boundary method with direct forcing for the simulation of particulate flows, J. Comput. Phys. 209 (2005) 448–476.]. The present paper proposes several enhancements of this method which considerably improve the results and extend the range of applicability. An important step is a simple low-cost iterative procedure for the Euler–Lagrange coupling yielding a substantially better imposition of boundary conditions at the interface, even for large time steps. Furthermore, it is known that the basic method is restricted to ratios of particle density and fluid density larger than some critical value above 1, hence excluding, for example, non-buoyant particles. This can be remedied by an efficient integration step for the artificial flow field inside the particles to extend the accessible density range down to 0.3. This paper also shows that the basic scheme is inconsistent when moving surfaces are allowed to approach closer than twice the step size. A remedy is developed based on excluding from the force computation all surface markers whose stencil overlaps with the stencil of a marker located on the surface of a collision partner. The resulting algorithm is throughly validated and is demonstrated to substantially improve upon the original method.

Coupling multibody dynamics and computational fluid dynamics on 8192 processor cores

This paper describes a method for the fully resolved simulation of particle laden flows. For this purpose, we discuss the parallelization of large scale coupled fluid structure interaction with up to 37 million geometrically modeled moving objects incorporated in the flow. The simulation is performed using a 3D lattice Boltzmann solver for the fluid flow and a so-called rigid body physics engine for the treatment of the objects. The numerical algorithms and the parallelization are discussed in detail. Furthermore, performance results are presented for test cases on up to 8192 processor cores running on an SGI Altix supercomputer. The approach enables a detailed simulation of large scale particulate flows that are relevant for many industrial applications.

Analytical assessment of models for large eddy simulation of particle laden flow

In Large-Eddy simulation of particle-laden flow, the effect of the unresolved scales on the particles needs to be modelled. In this work we analyse three very promising models, namely the approximate deconvolution method (ADM) which was proposed for particle-laden flow independently by Kuerten (Phys. Fluids 18, 2006) and Shotorban and Mashayek (Phys. Fluids 17, 2005) and two stochastic models, proposed by Shotorban and Mashayek (J. Turbul. 7, 2006) and Simonin et al. (Appl. Sci. Res. 51, 1993). We present results from a priori and a posteriori analysis of these models in isotropic turbulence at Re_lambda=52. This data allows for a direct quantitative comparison of the models. The analysis shows that ADM always leads to improved statistics but that even for high Stokes numbers, the rate of dispersion is not predicted correctly by ADM. Concerning the stochastic models, we found that with the correct choice of model parameters, the models perform well at small Stokes numbers. On the other hand, at high Stokes numbers the stochastic models show significant errors such that it may be recommendable to neglect the small scale effects instead of using one of the stochastic models.

Particle acceleration in turbulent flows: A class of nonlinear stochastic models for intermittency

The problem of modeling the velocity and acceleration of inertial particles in turbulent flows is discussed. Particular attention is focused on the modeling of the particle Lagrangian velocity increment, especially, but not exclusively, in the case in which only the low frequencies of the carrier turbulent flow field are available. The need for suitable models arises in the simulation of particle laden flows by the means of new computational techniques such as large-eddy simulation. For this, stochastic differential equations, sde, are often introduced, though there is a lack of clarity in how such models should deal with the experimental observed far from Gaussian statistics, intermittency, and heavy tailed probability density function for particle acceleration. It is well known that Langevin-type equations are not capable of reproducing such features. It is first shown how the stochastic model for the particle Lagrangian velocity increments is far from being a Langevin equation, and it is characterized by nonlinear drift and diffusion; the statistical characteristics of this first model are shown to be in qualitative agreement with experimental findings. These results suggest an improved model for the particle dynamics based upon a more general family of nonlinear sde; the family, which is generated by a single parameter, includes both the Langevin equation and the first model as special cases. An analysis of the statistical properties of the new sde shows that the model is capable of accurately reproducing the strong deviations from Gaussianity observed in recent experiments.