doi:10.1016/j.cma.2007.08.010
Copyright © 2007 Elsevier B.V. All rights reserved.
Coupling volume-of-fluid based interface reconstructions with the extended finite element method
John Dolbowa,
,
, Stewart Mossob, Joshua Robbinsb and Tom Vothb
aDepartment of Civil and Environmental Engineering, Duke University, Durham, NC 27708-0287, USA
bSandia National Laboratories, Albuquerque, NM 87185-0378, USA
1
Received 3 March 2007;
revised 2 July 2007;
accepted 4 August 2007.
Available online 7 September 2007.
References and further reading may be available for this article. To view references and further reading you must
purchase this article.
Abstract
We examine the coupling of the patterned-interface-reconstruction (PIR) algorithm with the extended finite element method (X-FEM) for general multi-material problems over structured and unstructured meshes. The coupled method offers the advantages of allowing for local, element-based reconstructions of the interface, and facilitates the imposition of discrete conservation laws. Of particular note is the use of an interface representation that is volume-of-fluid based, giving rise to a segmented interface representation that is not continuous across element boundaries. In conjunction with such a representation, we employ enrichment with the ridge function for treating material interfaces and an analog to Heaviside enrichment for treating free surfaces. We examine a series of benchmark problems that quantify the convergence aspects of the coupled method and examine the sensitivity to noise in the interface reconstruction. The fidelity of a remapping strategy is also examined for a moving interface problem.
Keywords: Interface reconstruction; Extended finite element; Volume-of-fluid
Fig. 2. Two-dimensional example of a bulk partition of the domain into a finite element mesh and corresponding partition of the interface into a set of element planes.
Fig. 3. Illustration of volume fraction gradient computation. The algorithm will calculate the gradient in the center element (indicated with heavy lines). The volume fractions will be averaged for each vertex using control volumes. The dashed lines indicate the boundary of the control volume for the top-most vertex. The edges of the control volume connect element centroids-to-centroids of element edges.
Fig. 4. Illustration of input configuration to planar smoothing using stability points. The dotted line is the given straight interface, and the solid lines in each element indicate the initial interfacial segments. The filled circles are the stability points, located at the centroid of each segment.
Fig. 5. Sample PIR reconstructed interfaces. Two-dimensional curved non-spherical test on an unstructured grid (left), and triple point, planar interfaces on an unstructured grid (right).
Fig. 6. Geometry and notation for one-sided benchmark problem.
Fig. 7. Coarse mesh (left) and zoom of interface (right) for one-sided problem.
Fig. 8. Numerical approximation uh to the solution u of the one-sided problem.
Fig. 9. Convergence in L2 error for the one-sided problem.
Fig. 10. Geometry and notation for two-sided benchmark problem.
Fig. 11. Coarse structured mesh and zoom near interface for two-sided problem.
Fig. 12. Convergence in energy norm for the two-sided problem. Angles listed in the legend indicate the degree of random perturbation given to the interfacial normal. Error bars are indicative of the range of results obtained using various random seeds in each case.
Fig. 13. Coarse unstructured mesh (left) and corresponding convergence results (right) for two-sided problem.
Fig. 14. Evolving interface problem: (left) mesh and interface at equally-spaced time steps; (right) interface segments only.
Fig. 15. Evolving interface geometry, at equally-spaced time steps on two subsequently refined meshes.