A parallel solution – adaptive method for three-dimensional turbulent non-premixed combusting flows

Gao, Xinfeng; Groth, C. P. T.

doi:10.1016/j.jcp.2010.01.001

Cited by 70 publications

(77 citation statements)

References 111 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The main distinction between different orders of accuracy is in the requirements on the number of ghost cell layers attached to each solution block. As in [15] and the previous work of Groth et al, e.g., [5,7,46,55,56], an efficient domain partitioning is achieved in our implementation by distributing the active solution blocks equally among available processor cores, with more than one block permitted per processor core. This approach efficiently exploits the self-similar nature of the solution blocks and readily produces an effective load balancing.…”

Section: Parallelization With Uniform Treatment Of Sector Boundaries mentioning

confidence: 93%

“…The 3D cubed-sphere grid simulation framework used in this work is based on a genuine multi-block implementation that was initially developed by Gao and Groth [45,46] for reacting flows and was later extended and optimized for cubed-sphere grids and space-physics flows by Ivan et al [14,15]. The computational framework allows for unstructured connectivity between root blocks, which is one of the primary nonstandard technical elements that has permitted the handling of cubed-sphere grids.…”

Section: Multi-block Cubed-sphere Grid With Unstructured Root-block Cmentioning

confidence: 99%

“…1(b). Furthermore, the computational framework allows for mesh adaptation accomplished by the dividing and coarsening of appropriate solution blocks, as described in [14,15,45,46]. The preliminary numerical tests in the current work do not take advantage yet of the full dynamic AMR capability of the computational framework, but use only the refinement feature to generate many solution blocks for large-scale parallel simulations.…”

Section: Multi-block Cubed-sphere Grid With Unstructured Root-block Cmentioning

confidence: 99%

“…The orientation of index axes in adjacent blocks (i.e., the orientation of i, j, and k indices in the logically Cartesian data structure of neighbouring blocks in physical space) is efficiently stored in compact form as a three-component transformation array. These transformation arrays provide a convenient short-hand notation for the transformation matrices [47,45,46] describing the relation between indices of two adjacent blocks, which can be used to exchange solution information between blocks having common interfaces in a general and transparent way. More details about the transformation arrays for the particular case of cubed-sphere grids are given in [15].…”

Section: Multi-block Cubed-sphere Grid With Unstructured Root-block Cmentioning

confidence: 99%

“…3): grid cells adjacent to one of the eight sector corners have only seven neighbouring cells (in 2D), while all other cells have 8 neighbours (these neighbours are used in stencils for solution reconstruction and flux calculation, see the next sections). This issue is dealt with in our approach by automatically detecting blocks with such corner cells, and by assigning "collapsed" corner ghost cells to those blocks sharing the relevant corner (as in [46]). In practice, this is implemented by marking these collapsed ghost cells, assigning them dummy values, and not including them in the stencils for reconstruction computation, so grid cells adjacent to sector corners employ reduced stencil sizes.…”

Section: Multi-block Cubed-sphere Grid With Unstructured Root-block Cmentioning

confidence: 99%

See 4 more Smart Citations

High-order central ENO finite-volume scheme for ideal MHD

Susanto

Ivan

Sterck

et al. 2013

Journal of Computational Physics

View full text Add to dashboard Cite

A high-order central essentially non-oscillatory (CENO) finite-volume scheme is developed for the compressible ideal magnetohydrodynamics (MHD) equations solved on threedimensional (3D) cubed-sphere grids. The proposed formulation is an extension to 3D geometries of a recent high-order MHD CENO scheme developed on two-dimensional (2D) grids. The main technical challenge in extending the 2D method to 3D cubed-sphere grids is to properly handle the nonplanar cell faces that arise in cubed-sphere grids. This difficulty is solved by considering general hexahedral cells with trilinear faces, which allow us to compute fluxes, areas and volumes with high-order accuracy by transforming to a reference cubic cell. The 3D CENO scheme is implemented within a flexible multi-block cubed-sphere grid framework to fourth-order accuracy, resulting in a high-order solution method for cubed-sphere grids with unique capabilities in terms of adaptive refinement and parallel scalability. The high-order method is applicable to the solution of general hyperbolic conservation laws with, in principle, arbitrary order. The CENO scheme is based on a hybrid solution reconstruction procedure that provides high-order accuracy in smooth regions, even for smooth extrema, and non-oscillatory transitions at discontinuities. The scheme is applied herein to MHD in combination with a GLM divergence correction technique to control the solenoidal condition for the magnetic field while preserving the high-order accuracy of the numerical procedure. The cubed-sphere simulation framework features a flexible design based on a genuine multi-block implementation, leading to high-order accuracy, flux calculation, adaptivity and parallelism that are fully transparent to the boundaries between the six sectors of the cubedsphere grid. The proposed 3D MHD CENO scheme is shown to achieve uniform fourth-order accuracy on cubed-sphere grids. Parallel domain partitioning and grid adaptivity are achieved on the 3D cubed-sphere grids using a hierarchical block-based division strategy with blocks of equal size. Numerical results to demonstrate the high-order accuracy, robustness and capability of the proposed high-order framework are presented and discussed for several test problems.

show abstract

Section: Parallelization With Uniform Treatment Of Sector Boundaries mentioning

confidence: 93%

Section: Multi-block Cubed-sphere Grid With Unstructured Root-block Cmentioning

confidence: 99%

Section: Multi-block Cubed-sphere Grid With Unstructured Root-block Cmentioning

confidence: 99%

Section: Multi-block Cubed-sphere Grid With Unstructured Root-block Cmentioning

confidence: 99%

Section: Multi-block Cubed-sphere Grid With Unstructured Root-block Cmentioning

confidence: 99%

See 3 more Smart Citations

High-order central ENO finite-volume scheme for ideal MHD

Susanto

Ivan

Sterck

et al. 2013

Journal of Computational Physics

View full text Add to dashboard Cite

show abstract

Hybrid mesh adaptation applied to industrial numerical combustion

Sirois

McKenty²,

Gravel³

et al. 2011

Numerical Methods in Fluids

View full text Add to dashboard Cite

SUMMARY This paper presents an anisotropic mesh adaptation method applied to industrial combustion problems. The method is based on a measure of the distance between two Riemannian metrics called metric non‐conformity. This measure, which can be used to build a cost function to adapt meshes comprising several types of mesh elements, provides the basis for a generic mesh adaptation approach applicable to various types of physical problems governed by partial differential equations. The approach is shown to be applicable to industrial combustion problems, through the specification of a target metric computed as the intersection of several Hessian matrices reconstructed from the main variables of the governing equations. Numerical results show that the approach is cost effective in that it can drastically improve the prediction of temperature and species distributions in the flame region of a combustor while reducing computational cost. The results can be used as a basis for pollutant prediction models. Copyright © 2011 John Wiley & Sons, Ltd.

show abstract

Accurate and robust adaptive mesh refinement for aerodynamic simulation with multi‐block structured curvilinear mesh

Chen

2015

Numerical Methods in Fluids

View full text Add to dashboard Cite

SUMMARYA multi-block curvilinear mesh-based adaptive mesh refinement (AMR) method is developed to satisfy the competing objectives of improving accuracy and reducing cost. Body-fitted curvilinear mesh-based AMR is used to capture flow details of various length scales. A series of efforts are made to guarantee the accuracy and robustness of the AMR system. A physics-based refinement function is proposed, which is proved to be able to detect both shock wave and vortical flow. The curvilinear mesh is refined with cubic interpolation, which guarantees the aspect ratio and smoothness. Furthermore, to enable its application in complex configurations, a sub-block-based refinement strategy is developed to avoid generating invalid mesh, which is the consequence of non-smooth mesh lines or singular geometry features. A newfound problem of smaller wall distance, which negatively affects the stability and is never reported in the literature, is also discussed in detail, and an improved strategy is proposed. Together with the high-accuracy numerical scheme, a multiblock curvilinear mesh-based AMR system is developed. With a series of test cases, the current method is verified to be accurate and robust and be able to automatically capture the flow details at great cost saving compared with the global refinement.

show abstract

A parallel solution – adaptive method for three-dimensional turbulent non-premixed combusting flows

Cited by 70 publications

References 111 publications

High-order central ENO finite-volume scheme for ideal MHD

High-order central ENO finite-volume scheme for ideal MHD

Hybrid mesh adaptation applied to industrial numerical combustion

Accurate and robust adaptive mesh refinement for aerodynamic simulation with multi‐block structured curvilinear mesh

Contact Info

Product

Resources

About