|
|
Simulating particle-laden flows: from immersed boundaries towards model order reduction
Isoz, Martin ; Kubíčková, Lucie ; Kotouč Šourek, M. ; Studeník, Ondřej ; Kovárnová, A.
Particle-laden flow is prevalent both in nature and in industry. Its appearance ranges from the trans-port of riverbed sediments towards the magma flow, from the deposition of catalytic material inside particulate matter filters in automotive exhaust gas aftertreatment towards the slurry transport in dredging operations. In this contribution, we focus on the particle-resolved direct numerical simulation (PR-DNS) of the particle-laden flow. Such a simulation combines the standard Eulerian approach to computational fluid dynamics (CFD) with inclusion of particles via a variant of the immersed boundary method (IBM) and tracking of the particles movement using a discrete element method (DEM). Provided the used DEM allows for collisions of arbitrarily shaped particles, PR-DNS is based (almost) entirely on first principles, and as such it is a truly high-fidelity model. The downside of PR-DNS is its immense computational cost. In this work, we focus on three possibilities of alleviating the computational cost of PR-DNS: (i) replacing PR-DNS by PR-LES or PR-RANS, while the latter requires combining IBM with wall functions, (ii) improving efficiency of DEM contact solution via adaptively refined virtual mesh, and (iii) developing a method of model order reduction specifically tailored to PR-DNS of particle-laden flows.
|
|
|
Octree-generated virtual mesh for improved contact resolution in CFD-Dem coupling
Studeník, O. ; Kotouč Šourek, M. ; Isoz, Martin
The present work is focused on improving the efficiency of a computational fluid dynamics (CFD) – discrete element method (DEM) solver allowing for computations with non-spherical solids. In general, the combination of CFD and DEM allows for simulations of freely moving solid particles within a computational domain containing fluid. The standard approach of CFD-DEM solvers is to approximate solid bodies by spheres, the geometry of which can be fully defined via its radius and center position. Consequently, the standard DEM contact models are based on an overlap depth between particles, which can be easily evaluated for a sphere-sphere contact. However, for a contact between two non-spherical particles, the overlap depth cannot be used and has to be replaced by the more general overlap volume. The precision of the overlap volume computation is (i) crucial for the correct evaluation of contact forces, and (ii) directly dependent on the computational mesh resolution. Still, the contact volume evaluation in DEM for arbitrarily shaped bodies is usually by at least one order of magnitude more demanding on the mesh resolution than the CFD. In order to improve the computational efficiency of our CFD-DEM solver, we introduce the concept of an OCTREEbased virtual mesh, in which the DEM spatial discretization is adaptively refined while the CFD mesh remains unchanged.
|
|
|
Hybrid fictitious domain-immersed boundary method in CFD-based topology optimization
Kubíčková, Lucie ; Isoz, Martin
Advances in technological development, especially in 3D printing, allow engineers to design components with almost arbitrary shape and connectivity. Consequently, more and more attention is being directed towards a highly-specialized application-driven component design based on topology optimization (TO). In the present work, we propose a methodology enabling TO of components in contact with flowing fluids. In particular, the optimization itself is based on multi-objective evolutionary algorithms (MOEAs) with the component geometry encoded using a binary representation. The optimization criteria are evaluated via computational fluid dynamics (CFD). The main novelty of the proposed TO framework lies in its robustness and effectiveness achieved by utilizing a single computational mesh for all the tested designs and projecting the specific components shapes onto it by the means of an immersed boundary method. The new methodology capabilities are illustrated on a shape optimization of a diffuser equipped as a part of an ejector. The optimization goal was to increase the ejector energy efficiency. The newly proposed methodology was able to identify a design by roughly 9 % more efficient than an alternative one found utilizing a previously published and less general optimization approach.
|
|
|
Shifted proper orthogonal decomposition and artificial neural networks for time-continuous reduced order models of transport-dominated systems
Kovárnová, A. ; Krah, P. ; Reiss, J. ; Isoz, Martin
Transport-dominated systems are pervasive in both industrial and scientific applications. However, they provide a challenge for common mode-based model order reduction (MOR) approaches, as they often require a large number of linear modes to obtain a sufficiently accurate reduced order model (ROM). In this work, we utilize the shifted proper orthogonal decomposition (sPOD), a methodology tailored for MOR of transport-dominated systems, and combine it with an interpolation based on artificial neural networks (ANN) to obtain a time-continuous ROM usable in engineering practice. The resulting MOR framework is purely data-driven, i.e., it does not require any information on the full order model (FOM) structure, which extends its applicability. On the other hand, compared to the standard projection-based approaches to MOR, the dimensionality reduction utilizing sPOD and ANN is significantly more computationally expensive since it requires a solution of high-dimensional optimization problems.
|
|
|
Estimating rheological properties of suspensions formed of arbitrarily-shaped particles via CDF-Dem
Kotouč Šourek, M. ; Isoz, Martin
In recent years, new methods combining computational fluid dynamics (CFD) and discrete element method (DEM) have been intensively studied. Usually, these methods are focused on simulations of spherical particles. Nevertheless, this is inadequate for a simulation of a common suspension, the rheology of which is affected by particle shapes. In this work, we leverage the capabilities of an in-house developed CFD-DEM solver to simulate suspensions formed of arbitrarily-shaped particles. Specifically, we simulate a rheological measurement to estimate the suspension viscosity. The CFD-DEM estimates are in very good agreement with available experimental data and correlations proving the new solver capabilities regarding firstprinciples-based simulations of complex non-Newtonian suspension behaviour. The practical potential of suspension simulation is illustrated in a numerical study of the washcoating process in the preparation of a catalytic filter for automotive exhaust gas after-treatment.
|
|
|
A parallel algorithm for flux-based bounded scalar Re-distribution
Isoz, Martin ; Plachá, M.
Let us assume a bounded scalar function ? : Q = I × ? ? ?0, 1?, I ? R, ? ? R3, where Q is an open bounded domain and its discrete counterpart ?h defined on a computational mesh Qh = Ih × ?h. The problem of redistribution of ?h over ?h ensuring the scalar boundedness while maintaining the invariance of R ?h ?h dV is surprisingly frequent within the field of computational fluid dynamics (CFD). The present contribution is motivated by the case arising from coupling Lagrangian particle tracking and particle deposition within ? h with Eulerian CFD computation. We propose an algorithm for ?h redistribution that is (i) based on fluxes over the computational cells faces, i.e. suitable for finite volume (FV) computations, (ii) localized, meaning that a cell ?h P with ?hP > 1 affects only its closest neighbors with ?h < 1, and (iii) designed for parallel computations leveraging the standard domain decomposition methods.
|
|
|
Developing a coupled CFD solver for mass, momentum and heat transport in catalytic filters
Hlavatý, Tomáš ; Isoz, Martin ; Kočí, P.
Using catalytic filters (CF) in automotive exhaust gas aftertreatment decreases the system heat losses and facilitates the CF regeneration. On the other hand, the CF overall performance is strongly dependent on the catalytic material distribution within it. In the present work, we aim to provide a computational framework to study the dependence of the CF characteristics, i.e. the pressure loss and the conversion of gaseous pollutants, on the catalyst distribution. Previously, we built an isothermal computational fluid dynamics (CFD) model of the flow and conversion of gaseous pollutants inside the CF. However, the reactions occurring inside the CF are exothermic and the assumption of constant temperature proved to be too restricting for real-life applications of the developed isothermal CFD model. Thus, in this work, we extend the framework by the enthalpy balance, which requires combining all the transport equations (mass, momentum and enthalpy) in a single solver. The new and more general solver provides results in good agreement with a well established (1+1)D channel model calibrated on experimental data. Furthermore, it allows studying more complex device-scale geometries of laboratory CF samples.
|
|
|
Recent improvements in CFD solver for fully coupled particle-laden flows
Šourek, M. ; Isoz, Martin
While new methods combining the computational fluid dynamics (CFD) and the discrete element method (DEM) have been developed to simulate freely moving solid particles, they tend to be focused on simulations with spherical particles only. Here, we present a strongly coupled CFD-DEM solver capable of simulating movement of arbitrarily-shaped particles dispersed in a fluid. The particles are assumed to be large enough to affect the fluid flow and distributed densely enough to come into contact with both the boundaries of the computational domain and with each other. In this paper, we will focus on the recent improvements of our solver, specifically, in the areas of (i) inclusion of solid bodies into the computational domain, (ii) general CFD-DEM coupling algorithm, and (iii) code parallelization and practical usability.
|
|
|
Increasing Ejector Efficiency via Diffuser Shape Optimization
Kubíčková, Lucie ; Isoz, Martin ; Haidl, Jan
An ejector is a technologically simple and yet wide-application fluid machine. While it has favorable characteristics for a signifficant number of technological processes, its main downside is probably its high operational energy demands. The present paper is an initial result of an ongoing research aimed at improving energy e ciency of the ejector via optimization of its geometry. In the paper, we focus mostly on presenting a general multi-objective optimization framework usable for an ejector shape optimization. The approach applicability is illustrated on a simpli ed problem comprising only a single phase flow in an ejector mixing tube and diffuser. Nevertheless, the achieved simulation and optimization results are validated against experimental data. The proposed optimization method itself is based on multi-objective evolutionary algorithms (MOEAs) combined with computational fluid dynamics (CFD) for evaluation of the vector-valued objective function.
|
|
|
Development of CFD solver for four-way coupled particle-laden flows
Šourek, M. ; Isoz, Martin
Computational uid dynamics (CFD) simulations containing freely moving bodies are still a challenging topic. More so, if the bodies are large enough to a_ect the uid ow and distributed\ndensely enough to come in contact both with the boundaries of the computational domain and with each other. In this work, we concentrate on the topic of simulation of (i) irregular bodies\nwith ow-induced movement and contact with computational domain boundaries taken into account, and (ii) bodies entrained by the uid and coming in contact not only with the domain\nboundaries but also with each other. The developed modeling approach is based on the hybrid _ctitious domain-immersed boundary method extended by the discrete element method. The\npresent contribution is focused on presentation of simulation principles and results of initial benchmark cases.
|
guest ::
