Digital Twins for metal additive manufacturing processes
For improved product quality and increased process productivity in metal additive manufacturing.
In the end-to-end digitalisation of manufacturing expected in an Industry 4.0 economy of the future, machines will be able to take action based on insights developed by artificial intelligence and advanced analytics using a wealth of real-time process data gathered by sensors. These cyber-physical actions will be aimed at reducing business costs by getting products and processes right every time. In order to succeed, however, process intelligence that is sufficiently in-depth is required. Such intelligence may be provided by physics-based computational models, which are called ‘digital twins’ of the processes. Advance knowledge derived from robust predictive capabilities of the digital twins, informed by accurate multiscale physical simulations, is especially helpful for the supervision of an additive manufacturing process which typically creates a high-value item that is a customised one-off. Such knowledge could help define an optimal processing window for a product that has not been manufactured before and assists with assuring quality in a part that is to be made for the very first time.
In a recent Symposium (July 2019) organised by CSIRO and held in Melbourne, Australia, hurdles to the development of the digital twins for additive manufacturing processes and their integration into a smart factory were discussed, along with potential solutions to these issues. Several international experts from diverse associated fields participated along with industry and government stakeholders. A summary of the Symposium may be found in an invited article published in the Fall 2019 edition of the Metal Additive Manufacturing magazine (page 185 onwards, vol. 5, no. 3) and available to download free at: [ddownload id=”929″]