Algorithmics & Software

Software tools and the algorithms they contain are an essential part of additive process optimization and the digital process chain. This includes dither methods for material positioning in multi-material 3D printing with minimal material agglomeration or geometry-adaptive laser guidance in laser beam melting, where the rapid calculation of various geometric features (distance to component surface, orientation of offset surfaces, etc.) is necessary. Methods for "making printable" 3D models with faulty geometry (e.g. incorrect surface orientation, holes, non-manifold edges and corners) are also essential for an automated process chain and fall into this category.

Models and Simulations

Simulation tools contribute to material development and to the optimization of process steps, e.g. to minimize distortions. The numerical approaches range from atomistic and thermodynamic simulations for alloy development to the flow, compacting and sintering behavior of powders, melt bath and microstructure development simulations of metals, mechanical models for the hardening behavior of resins, simulations of hybrid processes and post-treatment steps to the lifetime evaluation of additively manufactured components.

RISTRA - Rapid Interactive Structural Analysis

In the virtual product development process it is often crucial to simulate the properties of many product variants in order to generate resilient decision criteria for the selection of the final design prior to prototype construction or series production. This classical product simulation is frequently used in generative manufacturing, but also in other manufacturing processes. In order to significantly increase the number of variants to be investigated, we have developed RISTRA - Rapid Interactive Structural Analysis. RISTRA is a graphics processing unit (GPU) optimized simulation software for structural mechanics that uses thousands of GPU processors to quickly predict three-dimensional mechanical stress distribution under given loads. Through efficient GPU data structures and massively parallel algorithms, accelerations of up to 80x compared to commercial software are achieved.