August 29, 2023
Coreform Cubit, meshing software used to create high-quality meshes for challenging simulations, now has machine learning capabilities to speed up the meshing process, according to the company.
Coreform Cubit 2023.8 has been trained to recognize common fasteners such as bolts, nuts, pins and other connectors. Engineers can use this to identify these fasteners in their models and quickly simplify, remove or replace them. Users can also train Cubit to recognize their own common parts to perform similar simplifications. This machine learning technology has been developed by Sandia National Laboratory; the lead developer of the technology, Dr. Steven Owen, will give two free webinars to Coreform Cubit users, introducing the technology and how to use it.
The first webinar, Machine Learning in Coreform Cubit, Part 1: Part Defeaturing, will be held on August 31. The second webinar, Machine Learning in Coreform Cubit, Part 2: Part Classification and Reduction, will be held on September 7. Registration is now open for both webinars.
Additional functionality in Coreform Cubit 2023.7 includes improvements to the polyhedra meshing scheme, a new equi-angle skew metric, improved SDK documentation, user experience improvements and bug fixes. A complete list of these improvements is available in the Coreform Cubit documentation.
Additionally, the U-splines isogeometric analysis technology previously available in Cubit has been removed in advance of its inclusion in a new dedicated isogeometric analysis preprocessor being developed by Coreform. A new ‘.cf’ export file format has been added to Coreform Cubit to facilitate a future workflow.
The .cf format will be used to create body fitted meshes in Coreform Cubit for conversion to U-splines in a new software, Coreform Flex, which is currently in beta testing. Interested parties are invited to contact firstname.lastname@example.org if they wish to test this new IGA workflow.
Sources: Press materials received from the company and additional information gleaned from the company’s website.