MODSIM ​​​​​​​3DXCONFERENCE ​​​​​​​Lightweight Engineering 

Due to the underlying complex multi-physics systems, the spatiotemporal thermal cycles in additive manufacturing technology pose difficult challenges to validate the simulation models obtained from the finite element analysis (FEA). The spatiotemporal different thermal cycles lead to residual stresses and non-uniform part distortions within the additively manufactured part, which could detrimentally impact its adaptation in functional applications. Moreover, the conventional modeling using partial differential equations (PDEs) implemented through FEA is mitigated by the massive numerical calculations and exhausted by the system’s curse of dimensionality. In this study, a data-driven modeling approach was developed to simulate the distortion within the additively manufactured part effectively. The part distortions history was collected from different AM building strategies using finite element simulations. The FEA model consists of the multi-layer part distortion as targets for AM process parameters. The novel theme consists of two stages: (1) self-organized map (SOM) to project the high-dimensional spatial for the part distortion into likelihood estimator, (2) Hybrid self-organizing model to predict the extracted feature and reconstruct the part distortion field. The data-driven model evaluates different build strategies’ impact on the distortion for the additively manufactured part. The results correlated strongly with FE simulations and continued to establish a proper compensation strategy.

The replay:

The replay of this presentation will be available after the event via this post and I will be happy to answer your questions here as well.

@OA ​​​​​​​

You can find all event presentations on the wiki page: