InnoTop is an acronym for Interactive, Non-Linear, High-Resolution Topology Optimization, a project aimed at developing interactive and high resolution design tools for designing large scale constructions, while lowering their energy- and material consumption. InnoTop is a 6 years project running from 2017 to 2023, funded with 31 mill DKkr by Villum Foundation through a Villum Investigator grant awarded to Professor Ole Sigmund.
Topology Optimization is a wide-spread industrial method resulting in significant weight savings, reduced costs and lower energy consumption. Topology optimization is a numerical tool based on repeated finite element analysis, gradient evaluation and deterministic design updates used to determine optimal material distributions for mechanical and multi physic structures.
The InnoTop project will develop and expand algorithms to facilitate interactive, high-resolution topology optimization including shape morphing, stability, material nonlinearities, dynamics and manufacturing constraints.
A new high resolution, interactive design tool for large scale structures
InnoTop will create a high resolution interactive design tool, enabling design of large scale structures such as airplanes, bridges, wind turbines, ships and offshore structures, to the benefit of lighter, energy efficient and greener production and operation of structures.
Our challenge
The objective, an efficient high resolution design tool, is not achievable by simple algorithmic improvements or faster hardware alone. It requires breakthrough ideas and re-thinking of goals and parameterization approaches, as well as novel and advanced use of nonlinear and multi-scale procedures.
Our solution
Topology optimization was originally based on optimal microstructures through homogenization approaches, but this concept has almost entirely been substituted by simpler density and level-set approaches. Original concepts are revisited and combined with new ones, inspired by latest developments in computer graphics, multi-scale methods, non-linear homogenization approaches
and reduced order modelling.