Deep Neural Operator Enabled Digital Twin Modeling for Additive Manufacturing

Ning Liu,Xuxiao Li,Manoj R. Rajanna,Edward W. Reutzel,Brady Sawyer,Prahalada Rao,Jim Lua,Nam Phan,Yue Yu
2024-05-13
Abstract:A digital twin (DT), with the components of a physics-based model, a data-driven model, and a machine learning (ML) enabled efficient surrogate, behaves as a virtual twin of the real-world physical process. In terms of Laser Powder Bed Fusion (L-PBF) based additive manufacturing (AM), a DT can predict the current and future states of the melt pool and the resulting defects corresponding to the input laser parameters, evolve itself by assimilating in-situ sensor data, and optimize the laser parameters to mitigate defect formation. In this paper, we present a deep neural operator enabled computational framework of the DT for closed-loop feedback control of the L-PBF process. This is accomplished by building a high-fidelity computational model to accurately represent the melt pool states, an efficient surrogate model to approximate the melt pool solution field, followed by an physics-based procedure to extract information from the computed melt pool simulation that can further be correlated to the defect quantities of interest (e.g., surface roughness). In particular, we leverage the data generated from the high-fidelity physics-based model and train a series of Fourier neural operator (FNO) based ML models to effectively learn the relation between the input laser parameters and the corresponding full temperature field of the melt pool. Subsequently, a set of physics-informed variables such as the melt pool dimensions and the peak temperature can be extracted to compute the resulting defects. An optimization algorithm is then exercised to control laser input and minimize defects. On the other hand, the constructed DT can also evolve with the physical twin via offline finetuning and online material calibration. Finally, a probabilistic framework is adopted for uncertainty quantification. The developed DT is envisioned to guide the AM process and facilitate high-quality manufacturing.
Signal Processing,Artificial Intelligence
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to achieve closed - loop feedback control by constructing a digital twin (DT) model in the laser powder bed fusion (L - PBF) additive manufacturing process, in order to predict and optimize the molten pool state and defect formation. Specifically, the research aims to: 1. **Establish a high - fidelity physical model**: It is used to accurately represent the molten pool state and generate a data set for training machine - learning models. 2. **Develop an efficient surrogate model**: Utilize deep - learning techniques such as the Fourier neural operator (FNO) to quickly approximate the molten pool temperature field, thus supporting real - time monitoring and control. 3. **Extract physical information**: Extract information from the calculated molten pool simulations, such as molten pool size and peak temperature, which can be further related to the amount of defects (e.g., surface roughness). 4. **Optimize process parameters**: Optimize process parameters such as laser power and scanning speed by designing an objective function and using automatic differentiation techniques to reduce defects. 5. **Dynamically update the digital twin model**: Through offline fine - tuning and online material calibration, enable the digital twin model to evolve as the physical process progresses, for example, update the probability distribution of laser absorptivity to match the real - time captured thermal image data. 6. **Uncertainty quantification**: Adopt a probabilistic framework for uncertainty quantification to improve the robustness and reliability of the model. In summary, the goal of this research is to build a digital twin system that can predict and optimize the L - PBF process in real - time by integrating high - fidelity physical models, data - driven models, and machine - learning surrogate models, thereby improving manufacturing quality.