The SIAM International Meshing Roundtable Workshop 2022 is pleased to announce the following plenary talks to appear at this year’s conference.
- Prof. Jean-François Remacle
Feb. 23 - X-MESH: An eXtreme Mesh deformation method to follow sharp physical interfaces
- Dr. Adrien Loseille
Feb. 24 - Breaking the meshing pipeline with adaptivity
- Prof. Donna Calhoun
Feb. 25 - ForestClaw: A parallel library for solving PDEs on an adaptive hierarchy of logically Cartesian meshes
Prof. Jean-François Remacle
X-MESH: An eXtreme Mesh deformation method to follow sharp physical interfaces
Abstract: We develop an innovative approach, X-MESH, to overcome a major difficulty associated with engineering analysis: we aim to provide a revolutionary way to track physical interfaces in finite element simulations using extreme deformation of the meshes. Unprecedented low computational cost, high robustness and accuracy are expected as the proposed approach is designed to avoid the pitfalls of the current methods, especially for topological changes. The key idea of X-MESH is to allow elements to deform up to zero measure. For example, a triangle can deform to an edge or even a point. This idea is rather extreme and totally revisits the interaction between the meshing community and the computational community who, for decades, have striven to interact through beautiful meshes. Different areas in fluid and solid mechanics as well as heat transfer are targeted. Interfaces will be either (i) material, i.e. attached to particles of matter (the interface between two immiscible fluids or the dry interface in a wetting and drying model) (ii) immaterial, i.e. migrating through the material (a solidification front, contact front, yield front in yield stress fluid flow or a crack front). In this presentation, we will focus both on the mathematical issues related to the use of zero-measure elements and on the eXtreme mesh deformation scheme that will be used to track physical interfaces. Two applications will be targeted : phase change Stefan model and two phase flows.
Biography: After his Engineering Degree at the University of Liege in Belgium in 1992, Jean-François Remacle obtained in 1997 a Ph.D. from the same University. He then spent two years at the Ecole Polytechnique de Montréal as a post-doctoral fellow of Prof. F. Trochu, followed by three years at Rensselaer Polytechnic Institute in the research team of Prof. M. Shephard (one year as research associate followed by two years as research assistant professor). It was during his stay at Rensselaer that Pr. Remacle started to work closely with Mark Shephard on mesh generation. Pr. Shephard’s seminal work on mesh generation is one of the most important contributions ever. It was also during that stay that Pr. Remacle started the development of Gmsh, the open source mesh generator. After these five years in Northern America, Jean-François Remacle joined the Université catholique de Louvain in 2002 as an assistant Professor. He then became Associate Professor in 2005 and Full Professor in 2012. In the following years of his return to Europe, Pr. Remacle dedicated a large part of his research to mesh generation.
Since 2002, Pr. Remacle and his colleague Pr. C. Geuzaine from the University of Liège have continued the development of Gmsh. Gmsh was initially released as an open source in 2003 under the GNU General Public Licence (GPL). In 2009, a paper was published in the International Journal for Numerical Methods in Engineering (IJNME) that describes original features of Gmsh. This paper is the most cited paper of IJNME in the last 3 years. Gmsh was awarded a free software prize at the “trophées du libre” in 2009 . The size of Gmsh’s user community is now of over 8,000 regular users, including engineers of major European industries like Siemens, Dassault, EDF, Airbus or Snecma. Three Gmsh workshops have been organized, the last one in 2017.
In 2015, Pr. Remacle received a prestigious ERC Advanced Grant (www.hextreme.eu) with two major subjects: fast mesh generation and hexahedral mesh generation. Several breakthroughs have been achieved in HEXTREME, the three most significative ones being i) the developement of the fastest tetrahedral mesh generator, the use of Ginzburg Landau theory for generating quad meshes, and the development of an algorithm to build combinatorial hexahedral meshes whose boundary facets exactly match a given quadrangulation of the topological sphere.
Dr. Adrien Loseille
Breaking the meshing pipeline with adaptivity
Abstract: The mesh generation pipeline for scientific computing has been unchanged for decades. If many components have been studied on their own: from geometry, CAD preprocessing to surface and volume mesh generation, the process still relies on the following consecutive steps: geometry —> surface —> volume —> analysis. Human intervention is then generally needed to comply with the simulation/geometry/accuracy requirements and complexities. This increases design cycles while reducing the level of automaticity. The scope of this talk is to show how this pipeline is broken when adaptivity enters into the loop. In the context of numerical simulations, adaptivity has been seen as the art of “modifying” meshes to control numerical errors. Here, the goal is to show that adaptivity, when used within a proper theoretical framework, is more than simply controlling a level of accuracy. It can be used to define a robust by-design mesh generation and adaptation pipeline that allows to control geometry, element shapes and quality, and of course to allow naturally adaptivity in the loop. Several applications, from the aerospace industry, will illustrate this vision.
Biography: Adrien Loseille holds a PhD from Pierre et Marie Curie University, Paris VI, France and spent two years as a postdoctoral research associate at the CFD center at George Mason University. Following that, he was a full time researcher at INRIA, in the GAMMA group, and now works at Luminary Cloud. His core research is on adaptivity and unstructured adaptive meshing algorithms. He also has interests in high-order meshing and high-order visualisation. He is particularly interested in theoretically-sounded techniques while being applicable to complex industrial problems. The continuous mesh and the cavity operator framework are two examples that have allowed to leverage the state of the art of adaptive simulations. He has a long lasting collaboration with industry, his research has being supported by “The Boeing Company” or “Safran Teach”. His long term vision is to deliver the next generation meshing software. Now he takes on a new challenges with the development of mesh generation techniques in the cloud inside the tech start up LuminaryCloud.
Prof. Donna Calhoun
ForestClaw: A parallel library for solving PDEs on an adaptive hierarchy of logically Cartesian meshes
Abstract: We describe the software package ForestClaw, a PDE solver for time dependent PDEs based on updating a solution on an adaptive hierarchy of logically Cartesian meshes. The adaptive hierarchy is based the mesh generation library p4est (C. Burstedde, L. Wilcox and T. Isaac) for parallel dynamically adaptive mesh refinement (AMR) on a forest of octrees. ForestClaw is a PDE layer for p4est that provides time stepping, spatial discretizations, tagging criteria for refinement and all data transfer between neighboring grids and old and new meshes. I will describe the components of ForestClaw, the basic algorithms used to update the solution on a dynamically evolving quadtree mesh, and related issues of accuracy and stability of the AMR solution. I’ll describe several projects that are currently using ForestClaw, including solver strategies for elliptic problems. Recent results from a DARPA funded project to detected natural hazards (tsunamis, earthquakes, storms, volcanic eruptions) will also be discussed.
Biography: Donna Calhoun received her PhD in Applied Mathematics at the University of Washington (Seattle, WA). After post-docs at the Courant Institute (New York University, NYC) and Univ. of Washington, she spent almost six years working for the Atomic Energy Commission (CEA), in Saclay France. There, she developed simulations in support of the civilian nuclear industry in France. Following that, she joined the faculty in the Mathematics Department at Boise State University, where she teaches courses in mathematical and scientific computing, numerical methods, linear algebra and applied mathematics. Her research is supported by the National Science Foundation, NASA and DARPA.