Differentiable simulation for material thermal response design in additive manufacturing processes
Source abstract
Abstract
The flexibility of modern manufacturing processes such as additive manufacturing creates an opportunity to build parts with customized material properties and geometries. However, converting this flexibility into functionality requires computational tools that can handle high-dimensional design spaces. In this work, we present a differentiable simulation method for AM processes that is capable of designing time-series laser power that optimizes thermal behaviors of materials, including the overall thermal history and heat treatment time for each material point. We analyze the feasibility of computing gradients using automatic differentiation and their usefulness in AM design tasks. Additionally, our methodology is validated by experimentally testing our designed parts. Our code is made available for the research community at https://github.com/mojtabamozaffar/differentiable-simulation-am. © 2022 Elsevier B.V.
Source: openaire · Quality: candidate unverified · Source record ↗
Research context
OpenAlex topics: Additive Manufacturing and 3D Printing Technologies · Additive Manufacturing Materials and Processes · Manufacturing Process and Optimization
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References (39)
Show source-provided references
1. J.L. Bennett, K. Ehmann, J. Cao, Systems and methods for global thermal control of additive manufacturing, Google Patents, 2019.
2. Design guidelines for laser additive manufacturing of lightweight structures in TiAl6V410.2351/1.4885235 ↗
3. Design and development of an additive manufactured component by topology optimisation10.1016/j.procir.2017.03.027 ↗
4. Machine learning integrated design for additive manufacturing
5. Data-driven design of customized porous lattice sole fabricated by additive manufacturing10.1016/j.promfg.2021.06.035 ↗
6. Modeling, analysis, and optimization of dimensional accuracy of FDM-fabricated parts using definitive screening design and deep learning feedforward artificial neural network10.1007/s40436-020-00336-9 ↗
7. Online melt pool depth estimation during directed energy deposition using coaxial infrared camera, laser line scanner, and artificial neural network
8. Simulation-guided variable laser power design for melt pool depth control in directed energy deposition
9. Iterative simulation-based techniques for control of laser powder bed fusion additive manufacturing
10. Sensing, modeling and control for laser-based additive manufacturing10.1016/S0890-6955(02)00163-3 ↗
11. K.M. Taminger, R.A. Hafley, R.E. Martin, W.H. Hofmeister, Closed-loop process control for electron beam freeform fabrication and deposition processes, Google Patents, 2013.
12. Melt pool size control through multiple closed-loop modalities in laser-wire directed energy deposition of Ti-6Al-4V
13. Beyond the toolpath: site-specific melt pool size control enables printing of extra-toolpath geometry in laser wire-based directed energy deposition10.3390/app9204355 ↗
14. Site specific control of crystallographic grain orientation through electron beam additive manufacturing10.1179/1743284714Y.0000000734 ↗
15. Multi-energy source (MES) configuration for bead shape control in wire-based directed energy deposition (w-DED)10.1016/j.jmatprotec.2022.117549 ↗
16. End-to-end differentiable physics for learning and control
17. Iterative linear quadratic regulator design for nonlinear biological movement systems
18. Differentiable simulation for physical system identification10.1109/LRA.2021.3062323 ↗
19. DiffCloth: differentiable cloth simulation with dry frictional
20. Diffpd: differentiable projective dynamics with contact, arXiv e-prints
21. Improved protein structure prediction using potentials from deep learning10.1038/s41586-019-1923-7 ↗
22. Y. Hu, J. Liu, A. Spielberg, J.B. Tenenbaum, W.T. Freeman, J. Wu, D. Rus, W. Matusik, Chainqueen: A real-time differentiable physical simulator for soft robotics, 2019 International conference on robotics and automation (ICRA), IEEE, 2019, pp. 6265–6271.10.1109/ICRA.2019.8794333 ↗
23. J. Liang, M.C. Lin, Differentiable physics simulation, ICLR 2020 Workshop on Integration of Deep Neural Models and Differential Equations, 2020.
24. Interactive differentiable simulation
25. Scalable differentiable physics for learning and control
26. A. Paszke, S. Gross, S. Chintala, G. Chanan, E. Yang, Z. DeVito, Z. Lin, A. Desmaison, L. Antiga, A. Lerer, Automatic differentiation in pytorch, (2017).
27. M. Abadi, P. Barham, J. Chen, Z. Chen, A. Davis, J. Dean, M. Devin, S. Ghemawat, G. Irving, M. Isard, {TensorFlow}: A System for {Large-Scale} Machine Learning, 12th USENIX symposium on operating systems design and implementation (OSDI 16), 2016, pp. 265–283.
28. Difftaichi: differentiable programming for physical simulation
29. J. Bradbury, R. Frostig, P. Hawkins, M.J. Johnson, C. Leary, D. Maclaurin, S. Wanderman-Milne, JAX: composable transformations of Python+ NumPy programs, Version 0.1 55 (2018).
30. Jax md: a framework for differentiable physics
31. E. Heiden, D. Millard, E. Coumans, Y. Sheng, G.S. Sukhatme, NeuralSim: Augmenting differentiable simulators with neural networks, 2021 IEEE International Conference on Robotics and Automation (ICRA), IEEE, 2021, pp. 9474–9481.10.1109/ICRA48506.2021.9560935 ↗
32. P. Holl, V. Koltun, N. Thuerey, Learning to control pdes with differentiable physics, arXiv preprint arXiv:2001.07457 (2020).
33. Thermodynamically consistent microstructure prediction of additively manufactured materials10.1007/s00466-015-1243-1 ↗
34. Taylor expansion of the accumulated rounding error10.1007/BF01931367 ↗
35. J. Bergstra, O. Breuleux, F. Bastien, P. Lamblin, R. Pascanu, G. Desjardins, J. Turian, D. Warde-Farley, Y. Bengio, Theano: a CPU and GPU math expression compiler, Proceedings of the Python for scientific computing conference (SciPy), Austin, TX, 2010, pp. 1–7.10.25080/Majora-92bf1922-003 ↗
36. Mills · 2002
37. Adam: A method for stochastic optimization
38. Mechanical properties of hybrid additively manufactured Inconel 718 parts created via thermal control after secondary treatment processes10.1016/j.jmatprotec.2021.117047 ↗
39. Gradients are not all you need