Skip to main content
Log in

Parallel anisotropic 3D mesh adaptation by mesh modification

  • Original Article
  • Published:
Engineering with Computers Aims and scope Submit manuscript

Abstract

Improvements to a local modification-based anisotropic mesh adaptation procedure are presented. The first improvement focuses on control of the local operations that modify the mesh to satisfy the given anisotropic mesh metric field. The second is the parallelization of the mesh modification procedures to support effective parallel adaptive analysis. The resulting procedures are demonstrated on general curved 3D domains where the anisotropic mesh size field is defined by either an analytic expression or by an adaptive correction indicator as part of a flow solution process.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10
Fig. 11
Fig. 12

Similar content being viewed by others

Notes

  1. The basic unit to assign the destination partition id in the mesh migration procedure which is any mesh entity not on the boundary of any higher order entities.

  2. The scalability of a parallel program running on p processors against the program on one processor.

References

  1. Baker TJ (1997) Mesh adaptation strategies for problems in fluid dynamics. Finite Elem Anal Des 25(3–4):243–273

    Article  MATH  MathSciNet  Google Scholar 

  2. Apel T, Grosman S, Jimack PK, Meyer A (2004) A new methodology for anisotropic mesh refinement based upon error gradients. Appl Numerical Math 50(3–4):329–341

    Article  MATH  MathSciNet  Google Scholar 

  3. Bottasso CL (2004) Anisotropic mesh adaptation by metric-driven optimization. Int J Numer Meth Eng 60(3):597–639

    Article  MATH  MathSciNet  Google Scholar 

  4. Frey PJ, Alauzet F (2004) Anisotropic mesh adaptation for CFD simulation. Comput Methods Appl Mech Eng (accepted)

  5. George PL, Hecht F (1999) Non isotropic grids. In: Thompson J, Soni BK, Weatherill NP (eds) CRC Handbook of Grid Generation. CRC Press Inc., Boca Raton 20.1–20.29

    Google Scholar 

  6. Kunert G (2002) Toward anisotropic mesh construction and error estimation in the finite element method. Numerical Meth Partial Differential Equations 18:625–648

    Article  MATH  MathSciNet  Google Scholar 

  7. Li X, Shephard MS, Beall MW (2005) 3D anisotropic mesh adaptation using mesh modifications. Comput Methods Appl Mech Eng (in press)

  8. Pain CC, Umpleby AP, de Oliveira CRE, Goddard AJH (2001) Tetrahedral mesh optimisation and adaptivity for steady-state and transient finite element calculations. Comput Methods Appl Mech Eng 190:3771–3796

    Article  MATH  Google Scholar 

  9. Peraire J, Peiro J, Morgan K (1992) Adaptive remeshing for three dimensional compressible flow computation. J Comp Phys 103:269–285

    Article  MATH  Google Scholar 

  10. Remacle JF, Li X, Shephard MS, Flaherty JE (2003) Anisotropic adaptive simulation of transient flows using discontinuous Galerkin methods. Int J Num Meth Eng (in press)

  11. Tchon KF, Khachan M, Guibault F, Camarero R (2005) Three-dimensional anisotropic geometric metrics based on local domain curvature and thickness. CAD Comput Aided Des 37(2):173–187

    Google Scholar 

  12. Yamakaw S, Shimada K (2000) High quality anisotropic mesh generation via ellipsoial bubble packing. 9th International Meshing Roundtable

  13. Li X, Shephard MS, Beall MW (2003) Accounting for curved domains in mesh adaptation. Int J Numer Meth Eng 58:247–276

    Article  MATH  Google Scholar 

  14. de Cougny HL, Shephard MS (1999) Parallel refinement and coarsening of tetrahedral meshes. Int J Numer Meth Eng 46:1101–1125

    Article  MATH  Google Scholar 

  15. Seol ES, Shephard MS (2005) Efficient distributed mesh data structure for parallel automated adaptive analysis. Eng Comput (submitted)

  16. Seol ES, Shephard MS, Musser DR (2005) FMDB: Flexible distributed mesh database. http://www.scorec.rpi.edu/FMDB

  17. Frey PJ, George PL (2000) Mesh generation. Application to finite elements. Hermès Science Publishing, Paris, Oxford

    MATH  Google Scholar 

  18. Beall MW, Shephard MS (1997) A general topology-based mesh data structure. Int J Numer Meth Eng 40:1573–1596

    Article  MathSciNet  Google Scholar 

  19. Remacle JF, Klaas O, Flaherty JE, Shephard MS (2002) A parallel algorithm oriented mesh database. Eng Comput 18:274–284

    Article  Google Scholar 

  20. Gropp B et al (2005) The Message passing interface (MPI) standard library. Argonne National Laboratory http://www-unix.mcs.anl.gov/mpi

  21. Loy R (200) Autopack user manual. Science Division, Argonne National Laboratory

  22. Boman E, Devine KD, Hendrickson B et al (2001) Zoltan: a dynamic load-balancing library for parallel applications, Sandia National Labs, Zoltan’s user guide v1.23. http://www.cs.sandia.gov/zoltan

  23. Remacle JF, Flaherty JE, Shephard MS (2003) An adaptive discontinuous Galerkin technique with an othogonal basis applied to compressible flow problems. SIAM Rev 45:55–73

    Article  MathSciNet  Google Scholar 

  24. Barth TJ, Jespersen DC (1989) The design and application of upwind schemes on unstructured meshes. AIAA Paper 89-0366

  25. Krivodonova L, Xin J, Remacle JF, Chevaugeon N, Flaherty JE (2004) Shock detection and limiting with discontinuous Galerkin methods for hyperbolic conservation laws. Appl Numer Math 48:323–338

    Article  MathSciNet  MATH  Google Scholar 

Download references

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to E. Seegyoung Seol.

Appendix: Analytical size field definition

Appendix: Analytical size field definition

In this section, the four analytical size fields used in Sect. 4 are described.

A planar shock

We consider the following analytical size

$$ h_{1} = h_{{\rm max}} |1 - \hbox{e}^{-|x-0.5|}| + 0.003$$

with h max=0.2 and the analytical metric is given by:

$${\mathcal{M}} = {\mathcal{R}} \Lambda {\mathcal{R}}^{-1},\quad\hbox{with}\; \Lambda = \left(\begin{array}{*{20}c} h_{1}^{-2} & 0 & 0 \\ 0 & h_{{\rm max}}^{-2}&0 \\ 0 & 0 & h_{{\rm max}}^{-2}\\ \end{array} \right)\quad\hbox{and}\; {\mathcal{R}} = \left(\begin{array}{*{20}c} 1 & 0 & 0 \\ 0 & 1 & 0 \\ 0 & 0 & 1 \\ \end{array} \right). $$

A cylindrical shock

The analytical size field representing a cylindrical shock of radius R=1 and axis z=0 is defined by:

$$h_{1} = h_{{\rm max}} |1 - \hbox{e}^{-2 |x^{2} + y^{2} - R^{2}|}| + 0.0015$$

with h max=0.14 and the analytical metric is given by:

$${\mathcal{M}} = {\mathcal{R}}\Lambda {\mathcal{R}}^{-1},\quad\hbox{with}\;\Lambda = \left(\begin{array}{*{20}c} h_{1}^{-2} & 0 & 0 \\ 0 & h_{max}^{-2}&0 \\ 0 & 0 & h_{max}^{-2}\\ \end{array}\right)\quad\hbox{and}\; {\mathcal{R}} = \left(\begin{array}{*{20}c} x/r & -y/r & 0 \\ y/r & x/r & 0 \\ 0 & 0 & 1\\ \end{array} \right) .$$

where \(r=\sqrt{x^{2} + y^{2}}.\)

Two spherical shocks

We consider an analytical size field representing two spherical shocks of radius R=1 and centers (0, 0, 0) and (1, 0, 0), defined by:

$$\begin{aligned} h_{1} =& h_{{\rm max}} |1 - \hbox{e}^{-3|x^{2} + y^{2} + z^{2} - R^{2}|}| + 0.0015,\\ h_{2} =& h_{{\rm max}} |1 - \hbox{e}^{-3|(x-1)^{2} + y^{2} + z^{2} - R^{2}|}| + 0.0015\\ \end{aligned}$$

with h max=0.14 and the analytical metric is given by:

$${\mathcal{M}} = {\mathcal{R}} \Lambda {\mathcal{R}}^{-1},\quad \hbox{with}\; \Lambda = \left(\begin{array}{*{20}c} h_{1}^{-2} & 0 & 0 \\ 0 & h_{2}^{-2} & 0 \\ 0 & 0 & h_{max}^{-2}\\ \end{array} \right)\quad\hbox{and}\; {\mathcal{R}} = \left(\begin{array}{*{20}c} \vdots & \vdots & \vdots\\ v_{1} & v_{2} &v_{3}\\ \vdots & \vdots & \vdots\\ \end{array}\right), $$

where v 1=(x/r 1, y/r 1, z/r 1) with \(r_{1} = \sqrt{x^{2} + y^{2} + z^{2}},\) v 2 =((x−1)/r 2, y/r 2, z/r 2) with \(r_{2} = \sqrt{(x-1)^{2} + y^{2} + z^{2}}\) and v 3=v 1v 2.

In parallel, we define the same metric field with the following size:

$$\begin{aligned} h_{1} =& h_{{\rm max}} |1 - \hbox{e}^{-3|x^{2} + y^{2} + z^{2} - R^{2}|}| + 0.004 ,\\ h_{2} =& h_{{\rm max}} |1 - \hbox{e}^{-3|(x-1)^{2} + y^{2} + z^{2} - R^{2}|}| + 0.001\\ \end{aligned}$$

with h max=0.14.

Rights and permissions

Reprints and permissions

About this article

Cite this article

Alauzet, F., Li, X., Seol, E.S. et al. Parallel anisotropic 3D mesh adaptation by mesh modification. Engineering with Computers 21, 247–258 (2006). https://doi.org/10.1007/s00366-005-0009-3

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00366-005-0009-3

Keywords

Navigation