Authors:

Oliver Schenk (RWTH Aachen University, Germany),
Yuanbin Deng (RWTH Aachen University, Germany),
Christoph Broeckmann (RWTH Aachen University, Germany)

Abstract:

A digital twin offers the potential of understanding and improving production processes by using numerical models. Although multiple approaches of digital twins were reported, most of them only focus on the macroscale. However, the microstructural evolution is often crucial for the application of products. For instance, the pore morphology which determines the final mechanical properties is highly affected by the compaction and sintering steps in the PM process chain. In this work, a digital twin was developed to model the compaction and sintering of water-atomized Astaloy 85Mo on a mesoscale. Scanning electron microscopy images of green body microstructures were used to train a generative adversarial network to predict the microstructure dependent on the powder particle size. The evolution of these artificial microstructures during sintering due to surface diffusion was subsequently simulated with the level-set-method. The obtained sintering kinetic agrees well with that calculated by analytical equations of sinter neck growth.

DOI:

https://doi.org/10.59499/WP225371655