The use of overset meshes in particle and porous media three-dimensional flows

Dwyer, H. A.; Debus, Kristian; Shahcheraghi, N.

doi:10.1002/(sici)1097-0363(19990915)31:1<393::aid-fld985>3.0.co;2-i

Cited by 4 publications

(1 citation statement)

References 10 publications

(12 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The hybrid/overset/chimera/overlapping framework was firstly proposed by Steger, [8][9][10] which is advantageous to avoid meshing-topology changing when the objects move in the fluid domain. In this framework, there are two sets of meshes, Eulerian and Lagrangian.…”

Section: Introductionmentioning

confidence: 99%

A novel dynamic adaptive unstructured mesh algorithm for simulating multi‐object relative motion in incompressible fluid

Zhang

Zhao

et al. 2022

Numerical Methods in Fluids

View full text Add to dashboard Cite

A novel dynamic adaptive unstructured mesh (DAUM) algorithm is proposed to solve incompressible multi-object relative motion (MORM). The DAUM algorithm, consisting of exponential function deformation, adaptive edge swapping, and area Laplace smoothing, can greatly improve dynamic mesh robustness and perfectly overcome mesh skewness. The core of DAUM is the adaptive edge swapping inspired by Delaunay triangulation, which is distinguished from traditional edge swapping. The adaptive edge swapping can fully consider the relationship of neighbor elements only using Delaunay triangulation. Meanwhile, the implementation and reliability of adaptive edge swapping are better than the traditional edge swapping method due to eliminating the interference of the non-convex polygon, so more code remedies can be avoided. Using the DAUM, none of the vertices is inserted or deleted so that the manipulation of the dynamic mesh is easily implemented and maintains computational efficiency in the process of mesh motion. Three representative geometries are used to assess the performance of the DAUM in MORM. To systematically analyze the advantages of DAUM, two relatively moving cylinders have been numerically investigated in incompressible flow. The inline force and lift force profiles on two cylinders are obtained and analyzed by using the flow field information.Three interaction stages are divided based on the parameter G and the interactional intensity of two inner anticlockwise vortices is considered as the division criteria. At the running process of the DAUM algorithm, the dynamic mesh quality is well controlled and remains in the high-quality range based on the aspect ratio (AR) criterion. The results indicate that the proposed DAUM algorithm can properly solve the difficulties caused by MORM, especially for period oscillation motion.

show abstract

Section: Introductionmentioning

confidence: 99%

A novel dynamic adaptive unstructured mesh algorithm for simulating multi‐object relative motion in incompressible fluid

Zhang

Zhao

et al. 2022

Numerical Methods in Fluids

View full text Add to dashboard Cite

show abstract

Network modeling of flow in a packed bed

et al. 2006

View full text Add to dashboard Cite

in Wiley InterScience (www.interscience.wiley.com).The characterization of the flow inside a packed bed requires a full description of the pore geometrical characteristics, and of the flow phenomena at local level. A twodimensional (2-D) network model for describing the flow phenomena in unconsolidated-packed beds has been developed. The network model consists of two different types of elements: chambers modeled as spheres, and channels modeled as cylinders. The size distributions of the network elements are obtained considering a geometrical model that uses the porosity, and the average-particle diameter as input data. A flow simulator was developed, based on this network model. Results show that since the inertial effects due to connections between channels and chambers are taken into account, this simulator is capable of describing single-phase flow in all the possible flow regimes, from laminar to turbulent. Results also show a good agreement between predicted values of the network model and experimental data available in literature.

show abstract