A new addition to the Roundtable for 1999 is the optional short courses, to be held the day before the opening of the Conference. Three courses will be offered, taught by internationally known experts in the field of Mesh Generation. Each short course will be two hours in length and include course notes and coffee breaks. Instructors will be addressing practical issues in the design and implementation of both structured and unstructured mesh generation codes. The courses are ideal for students just entering the field needing a foundation for research, or for seasoned professionals who would like to expand their current skill-set in the development of mesh and grid generation algorithms. To register for the short courses, mark the appropriate boxes on the registration form. The price is $100 per attendee which includes course materials. Enrollment for short courses is limited to the first 30 paid registrants. If insufficient enrollment, short courses may be subject to cancellation. In event of cancellation or over-enrollment, all tuition will be refunded.
10:00-12:00 Basics of Structured Grid Generation
1:00 - 3:00 New Advances and Open Issues in Unstructured Meshing
3:30 - 5:30 Parallel Meshing and CAD Interface
During the last few years, computational field simulation(CFS) has emerged as an essential analysis tool for engineering design problems. An essential element of the CFS process for general region is the construction of a discrete grid on which the field equations are represented in finite form. This discrete representation can be classified as cartesian, structured (curvilinear and boundary conforming), unstructured, and hybrid/generalized grid.
The structured grid is presented by a network of curvilinear coordinate lines such that one-to-one mapping can be established between physical and computational space (uniformly spaced rectangular space). The non-linear partial differential equations representative of the field under study are solved in the computational space using efficient finite difference/finite volume techniques.
This session is designed for CFS practitioners to gain thorough understanding of basic foundation, strategies, generation methodologies and computational tools for multiblock structured grids. A step-by-step procedure associated with practical grid configurations will be developed. The topics will include: physical to computational space mappings involving various zonal/block strategies, transformation relations, truncation errors and grid quality and distribution requirements, geometry preparations, algebraic, elliptic, and hyperbolic generation methods, and grid adaptation. The computational algorithms and associated public and commercially available software will be discussed along with current bottlenecks, shortcomings and future trends.
Finite element analysis requires the discretization of the geometric model into a mesh; this mesh is often required to be composed of hexahedra. Producing a hexahedral mesh for component assemblies is a challenging task requiring the application of a number of different tools, including geometric decomposition, meshing algorithm selection, and others. This short course will survey the current algorithms and tools which make up the decomposition-based approach to hexahedral meshing. Information will be given on how each algorithm works, as well as how they can be combined to make an effective hex meshing process. Promising technologies still in the research stage will also be enumerated and described.
This lecture will address two topic areas of importance to the development of effective mesh generators for large-scale problems.
Since the fundamental descriptions of geometrically complex engineered artifacts are in terms of non-manifold solid models, the ability to effectively generate computational meshes for these artifacts is strongly influenced by our ability to work directly with the solid model representations. This discussion will first address basic issues of non-manifold topological model representations and the importance of maintaining the associativity of the mesh and the model. The influence of geometric modeling tolerances will also be discussed and a method to employ the functionality of the solid modeling kernel to support mesh generation will be discussed. Finally some issues and ideas with respect to dealing with "messy" or overly detailed geometric models will be discussed.
The discussion of parallel mesh generation will overview the
approaches taken that use message passing as the means to control
the parallel operations. In addition to indicating why it is difficult
to effectively generate meshes in parallel, a specific algorithm
will be explained in more detail.