CASOPT Consortium

Logo ABBLogo TU GrazLogo TUMLogo University of Cambridge

CASOPT Scientific Objective: Create a new methodology and algorithms for a novel, uniquely integrated Component-Level and Assembly-Level Optimization governed by physics-driven design criteria.
Using physics-driven design criteria as a control link between novel Component-Level and Assembly-Level Optimization technologies presents an entirely new approach in the future design of industrial apparatus in general. This scientific leap is the key to achieving the technological objectives.
 dive - physics-based

 

CASOPT Technological Objective: Maximize the operational device performance, such as device voltage or current level, and simultaneously minimize the geometric device dimensions to obtain the most efficient and compact design.


Above objectives will be achieved by launching a novel multi-physics, multi-level simulation platform that will require leveraging of the following technologies and know-how available at the partner sites:

  • Computational techniques for Component-Level Optimization with novel non-parametric / non-gradient algorithms
  • A cutting-edge Boundary Element Method (BEM) simulation framework for Laplace, Helmholtz, Kirchhoff and non-linear multi-physics problems (established and proven for industrial design applications)
  • Advanced heuristic and non-heuristic techniques for Assembly-Level Optimization
  • Optimized parallelization algorithms for efficient use of multi-core based architecture 
  • Computational geometry, meshing and related algorithms for smoothing, collapsing, healing and mesh adaptation
  • Software for converting simulation meshes featuring the optimal forms back to  CAD systems
  • Integrated simulation and geometric modeling concept based on new subdivision surfaces techniques
  • Industrial expertise on the design of geometrically and physically complex device assemblies

CASOPT: Controlled Component- and Assembly-Level Optimization of Industrial Devices | zoran.andjelic@ch.abb.com
SOLUTION·CH