(364an) Process Mapping and Optimization through Adaptive Machine Learning for a Laser Powder Bed Fusion (L-PBF) Additive Manufacturing Process

As machine learning technologies have emerged as powerful and useful tools, many innovators have been investigating how to use these tools to improve manufacturing/industrial processes. This has contributed to the emergence of another revolution in the area of industrial processes, known as Industry 4.0 [1]. In this work, we investigate how machine learning can be combined with domain knowledge in a systems engineering approach to process improvement. In particular, we use adaptive sampling methods through machine learning to efficiently build process maps for processes of interest.

We apply this adaptive sampling approach in this work to a pulse wave laser powder bed fusion (L-PBF) process. L-PBF is a popular additive manufacturing (AM) technology in which metal parts are built from computer aided design models through layer-by-layer laser sintering of metal powder. While this AM method has advantages over other manufacturing processes, particularly for its efficiency and low waste, it is a complex process with several variables to optimize for robust performance [2], [3]. In this work, we are interested in mapping the transition mode region, a region of high variability but also high density/strength parts [4], [5]. We show how adaptive sampling methods can be used in conjunction with single scan track (SST) experiments to robustly identify this region in the parameter space for the L-PBF process. We then apply symbolic methods to the experimental data to find physically-meaningful models for the transition mode region.

Overall, we show how adaptive sampling or iterative design of experiments (DOE) methods can be used along with symbolic methods to robustly and efficiently find physically-meaningful process models for complex processes.

[1] U. M. Dilberoglu, B. Gharehpapagh, U. Yaman, and M. Dolen, “The Role of Additive Manufacturing in the Era of Industry 4.0,” in Procedia Manufacturing, Elsevier B.V., 2017, pp. 545–554.

[2] E. Toyserkani, D. Sarker, O. O. Ibhadode, F. Liravi, P. Russo, and K. Taherkhani, “Basics of Metal Additive Manufacturing,” in Metal Additive Manufacturing, Wiley, 2021, pp. 31–87.

[3] S. S. Razvi, S. Feng, A. Narayanan, Y.-T. T. Lee, and P. Witherell, “A Review of Machine Learning Applications in Additive Manufacturing,” in International design engineering technical conferences and computers and information in engineering conference, Aug. 2019.

[4] S. Patel and M. L. Vlasea, “Melting Mode Thresholds in Laser Powder Bed Fusion and their Application Towards Process Parameter Development,” 2019.

[5] S. Patel and M. Vlasea, “Melting modes in laser powder bed fusion,” Materialia (Oxf), vol. 9, Mar. 2020.

Research Interests

My research interests lie at the intersection of machine learning and process optimization, with a particular focus on advanced manufacturing processes such as Laser Powder Bed Fusion (L-PBF). I am interested in areas where I can leverage domain knowledge and machine learning techniques to help gain meaningful insight on complex processes.

Keywords/terms: Machine learning, advanced manufacturing, additive manufacturing, process optimization