You are here

The Hexahedralization Sensation

Award Information
Agency: Department of Defense
Branch: Air Force
Contract: FA8650-18-P-2256
Agency Tracking Number: F173-007-0229
Amount: $150,000.00
Phase: Phase I
Program: SBIR
Solicitation Topic Code: AF173-007
Solicitation Number: 2017.3
Timeline
Solicitation Year: 2017
Award Year: 2018
Award Start Date (Proposal Award Date): 2018-04-12
Award End Date (Contract End Date): 2019-04-12
Small Business Information
1270 North Fairfield Rd
Dayton, OH 45432
United States
DUNS: 004475216
HUBZone Owned: No
Woman Owned: No
Socially and Economically Disadvantaged: No
Principal Investigator
 Christopher Koehler
 (937) 426-8530
 ckoehler@utcdayton.com
Business Contact
 Joe Sciabica
Phone: (937) 426-2808
Email: jsciabica@utcdayton.com
Research Institution
N/A
Abstract

Automatic conversion of computer aided drafting (CAD) geometry to computational grids appropriate for high fidelity hypersonic analysis is needed. With respect to grid quality, the output must be purely hexahedral and C2 continuous, even at extraordinary vertices. To accomplish this, a pipeline of local and global CAD repair algorithms will feed into a feature-preserving quadrangulation module with optional adaptive surface refinement. Catmull-Clark subdivision surfaces will be iteratively aligned to the quadrilateral surface meshes to provide a lightweight parametric representation. Recent advances from the computer graphics community will be employed to assure C2 continuity at extraordinary vertices. These techniques map a C2 continuous polynomial onto the subdivision surface with a C2 continuous compact weight function. The result will be projected back into the UV space of the input CAD to enable geometric surface sensitivities. Proven quadrangulation algorithms will then be extended to 3D hexahedralization to enable the automatic generation of purely hexahedral volumetric grids. Similarly, 3D subdivision volumes will map hexahedra to continuous, lightweight parametric representations. An additional library will be created to discretize the parametric grids and push them directly into processor memory. Bypassing the hard drive in this way is essential for parallel scalability to billions of grid cells.

* Information listed above is at the time of submission. *

US Flag An Official Website of the United States Government