MG Design and Implementation of a New Efficient Mesh.pptxVIP

CAPSTONE: Providing Geometry, Meshing and Attribution for Physics-based Analysis and Design Dr. Saikat Dey NRL Washington DC CREATE-MG Project Manager February 1, 2010, NDIA CREATE Meeting, Arlington, VA OUTLINE Motivation, Strategic needs and Challenges CAPSTONE – the product Overview Users and Usecase Scenarios Capability Development Roadmap Current status, use-cases and testing Summary Motivation Goal Improve efficiency of DoD acquisition engineering by reducing time, cost and risks in research, development and sustainment of weapon systems Critical Hurdles Human Effort Calendar Time to Produce an Analyzable Representation (Model) of a Design or System Significantly more time is often spent in ‘preparing’ the input data needed by solvers than is used by the solvers to solve it. Approach Develop Next-Generation Computational Solvers Optimizers Insert More (Multi) Physics-Based Analysis Earlier in the Design-Cycle Geometry and Meshing Needs “Let no one ignorant of geometry enter” - Plato Geometry needs to be appropriate for analysis and meshing Valid Dimensionally correct (1-,2-,3-D or mixed-dimension, non-manifold) “Water-tight” (no gaps), non-self-intersecting Accurate Match a shape to a given tolerance Maintain the accuracy and rate of convergence of the solvers/code Meshing needs to be appropriate for physics and discretization What takes time and effort ? Geometry repair/clean-up De-featuring (geometry good for Physics A is not suitable for Physics B) Lack of automation and robustness in meshing (all-hex, complex boundary layers) Attribution, multi-component model preparation CREATE-MG: Mission Summary Rapid, Scalable and automated generation of analyzable representations (geometry, mesh, attribution data) for accurate and scalable physics-based solvers Develop Capability and Tools for: Enabling: Multi-physics based analyses earlier in the design process Rapid turnaround time and automation key to effective design optimization Generation an