Lab Home | Phone | Search
Center for Nonlinear Studies  Center for Nonlinear Studies
 Home 
 People 
 Current 
 Executive Committee 
 Postdocs 
 Visitors 
 Students 
 Research 
 Publications 
 Conferences 
 Workshops 
 Sponsorship 
 Talks 
 Seminars 
 Postdoc Seminars Archive 
 Quantum Lunch 
 Quantum Lunch Archive 
 P/T Colloquia 
 Archive 
 Ulam Scholar 
 
 Postdoc Nominations 
 Student Requests 
 Student Program 
 Visitor Requests 
 Description 
 Past Visitors 
 Services 
 General 
 
 History of CNLS 
 
 Maps, Directions 
 CNLS Office 
 T-Division 
 LANL 
 
Monday, February 01, 2010
10:00 AM - 11:00 AM
CNLS Conference Room (TA-3, Bldg 1690)

Seminar

A multiphysics CFD framework for fluid-solid interaction

Mark A. Christon
CTO Office, Dassault Systemes SIMULIA

Incompressible flows are some of the most frequently encountered in industrial applications and vary from pollutant and gas dispersal to chemically reacting flows, food processing, encapsulation, casting, mold filling, and vehicle aerodynamics. For these classes of problems, mesh resolution ranging from $106$ to $108$ grid points may be required to resolve important physical phenomena and spatial features in complex geometry. The algorithmic challenges involved in solving time-dependent, incompressible flow problems hinge upon the div-free constraint, efficient treatment of the concomitant pressure equation and scalable, parallel solution algorithms. Abaqus/CFD is an entirely new software framework that provides not only a capability for multiphysics simulation with the Abaqus computational mechanics codes, but also the ability to easily support multiple physics in a single framework, e.g., incompressible flow, compressible flow, heat-conduction, thermal radiation, etc. The software framework is 100\% parallel, provides parallel load-balancing/data-migration tools, and is designed to scale to to thousands of processors. The flow solvers in Abaqus/CFD relies on edge-based data structures to implement both FVM and FEM solution algorithms. The incompressible flow solver is based on an ALE formulation with a second-order incremental projection method. The incompressible flow solver uses a hybrid FVM/FEM method that circumvents the well-known LBB div-stability, provides local-conservation and monotonicity-preserving advection for transport quantities. A parallel, algebraic multigrid preconditioner in combination with a CG solver is used to solve the node-based pressure-Poission equation providing a scalable solution strategy for a broad range of industrial applications. This talk will provide a brief survey of the multiphysics CFD framework and associated technology components. The kinematics and master balance relations for the ALE formation will be presented with the key extensions for the approximate projection method. The discretization and solution procedure will be presented with a series of industrial calculations demonstrating the application of the multiphysics CFD framework.

Host: Mikhail Shashkov