CAE Geometry: Engineering Models That Support Integrated, Automated Simulation

Add to your list(s) Download to your calendar using vCal

• Dr Tom Cashman, TranscenData Europe
• Thursday 04 December 2014, 11:15-12:15
• Rainbow Seminar Room (SS03), Computer Laboratory.

If you have a question about this talk, please contact Advait Sarkar.

CAD (Computer-Aided Design) and CAE (Computer-Aided Engineering) models are built to meet different requirements and serve very different purposes, resulting in a significant and costly disconnect between CAD and CAE systems. We explore three longstanding barriers to CAD /CAE integration: the challenge of finding high-performance meshes, the missing link in both directions between design and analysis, and the fundamental limitations of today’s CAD models. We discuss the need for domain partitioning, in order to generate meshes that give the best possible results with the shortest possible computation time. We argue that the idealisations of a CAD model that are necessary for CAE quickly introduce a disparity between these models, which is usually resolved by manual processing. We also describe the theoretical reasons why CAD models almost universally contain gaps and inconsistencies. We address components of all three of these issues by introducing alternative geometry representations that we describe as “CAE geometry”, with the key characteristics that they are smooth, consistent and watertight descriptions. One possibility is realised by representing face geometry with a smooth curved triangle mesh, so that CAD model edges are able to define the exact join between two neighbouring faces. We describe the direct use of this representation in structural FEA , to obtain a deformed version of the geometry, and the automated conversion of such a representation back to a standard CAD model. An alternative watertight description is provided by computing the Medial Object of a part, making it possible to automatically generate domain blockings that are suitable for hex meshing. This allows changes to a design to be automatically reflected in the analysis mesh, while retaining the mesh’s efficiency and fidelity. We conclude that future CAE will be facilitated by tools that automate the bidirectional flow of data between the CAD and CAE disciplines, making it possible to pursue an increasingly integrated process.

This talk is part of the Rainbow Group Seminars series.

This talk is included in these lists:

• All Talks (aka the CURE list)
• Cambridge Centre for Data-Driven Discovery (C2D3)
• Cambridge talks
• Chris Davis' list
• Department of Computer Science and Technology talks and seminars
• Interested Talks
• Rainbow Group Seminars
• Rainbow Seminar Room (SS03), Computer Laboratory
• School of Technology
• Trust & Technology Initiative - interesting events
• bld31
• ndk22's list
• ob366-ai4er
• rp587

Note that ex-directory lists are not shown.