Developing a Digital Twin for Metal 3D Printing

Metal Additive Manufacturing (MAM), particularly the Laser Powder Bed Fusion (L-PBF) process, is a frontier of modern engineering, offering unprecedented flexibility in designing and producing complex parts. To optimize these processes and ensure the highest quality, we've embarked on developing an intelligent digital twin based on a physics-based model that simulates the real-world manufacturing conditions. 

The goal?  

By leveraging both experimental data and simulation software, Twin4Twin aims to develop a predictive digital twin that reflects the real manufacturing process as closely as possible, providing valuable insights and predictive capabilities for enhancing the precision, efficiency, and quality of 3D metal printed components.  

In the context of Twin4Twin project the Laser Powder Bed Fusion (L-PBF) process has been selected as benchmark use-case, whereas the 3d printer by Aconity3D placed in the SSF facilities has been used as a test-bed for the experimental trials.  

Highlights of The Approach

First, a specific Metal AM scenario that reflects typical industry challenges has been selected that uses the L-PBF process, common in high-precision manufacturing fields such as aerospace and automotive.

Next step was to define the technical specifications and find a suitable geometry that will serve as the basis for both simulations and the real experimental trials.

To simulate the real process a general-purpose software, such as ABAQUS has been used, that is providing a robust environment for analyzing the complex dynamics involved. Moreover, AMPHYON that is a specifically designed software for Metal AM has been utilized to analyze stress formations and other critical factors of the real process.     

A solid framework for model validation has been utilized that uses both data from in-situ sensors and ex-situ measurements after the printing job has finished. After each printing job, the physical properties of the printed parts are analyzed and compared with model’s predictions.  

Figure 1: Validation workflow for Digital Twin

What's Next?

As next steps the Twin4Twin team aims to:

·proceed with the digital twin integration

·use advanced data analytics and machine learning to improve the digital twin accuracy

·incorporate more complex geometries, a wider range of materials and work in parallel with industry partners to validate and refine the digital twin in actual manufacturing settings

Stay tuned as Twin4Twin will continue to refine this technology and explore the full potential of digital twin models in complex manufacturing processes.

Partners’ Roles

CORE IC is leading the research work and is responsible for the integration of the Digital Twin;  ITA is developing the physics-based models simulating the real process; SSF provides the technical specifications for the test-bed and runs the full cycle of experimental trials in the real 3d printer; SCCH is responsible for the Big Data technologies and the IoT architecture.