(419f) Modeling Impacts of Cyberattacks on Control of Powder Bed Fusion

Next-generation manufacturing will involve new and innovative methods of manufacturing, requiring increased flexibility. Additive manufacturing methods (i.e., 3D printing) are enticing as they can handle many different configurations for producing unique and complex structures, including for prototyping and manufacturing, in a wide variety of fields [1, 2]. Powder bed fusion (PBF) is one kind of additive manufacturing method that can be used to produce metal objects through selectively melting powder using a laser. Between each layer, a roller or sweeping mechanism spreads a fresh layer of powder over the previously melted layer. The laser then fuses the next layer to the previous, and successive layers are added to create the final completed part. The advantage of PBF is that high-quality metal parts can be produced, but post-processing is typically resource and labor-intensive. This includes ensuring part quality, finishing, and heat treatment [3]. Current methods of improving part quality (and thus minimizing post-processing) usually involve tuning production parameters based on previous experience or testing [3, 4], or though control of the laser or laser melt pool [5]. Advanced control methods have been used for tracking a desired melt pool behavior [5]. However, as control concepts for additive manufacturing become more sophisticated, it becomes necessary to consider how attacks on control systems might impact additively manufactured parts. It has been recognized that a potential challenge for additive manufacturing is cyberattacks on the control systems [6], which could, for example, introduce defects in a part.

In [7, 8], we discussed simulation of rogue input policies in the context of a steam methane reformer (SMR) where both transport and equipment stress considerations were important; it would be expected that similar studies could aid in showcasing the impacts of attacks on additively manufactured parts. While both processes can be simulated in ANSYS with finite element methods, they consider different phenomena that require unique considerations. In particular, PBF must consider the evolving part shape which could be manipulated in many ways, including void creation, incomplete melting, warping or cracking, or altered dimensions.

This talk will build from our prior studies using ANSYS for simulating rogue control policies in the context of the more traditional chemical process of steam methane reforming to focus on additive manufacturing, and the criticality of understanding the effects of rogue control actions on additively manufactured parts. We will discuss our progress on considering cybersecurity within an additive manufacturing process simulation [9] where both transport phenomena (in this case, heat transfer) and stress in a part are important. We will discuss modeling of the powder bed fusion process with a single powder layer on a baseplate, where the process is modeled with finite element analysis (FEA) in ANSYS based on modeling strategies synthesized from [10-13]. The process was simulated using a three-dimensional geometry and through a coupled ANSYS thermal and structural analysis. ANSYS parametric design language (APDL) code was written to apply volumetric heat generation (based on [12]) to represent the temperature change induced by the laser, to manage the phases involved (powder, solid, and liquid), and to manage the coupling of the thermal and structural simulations. This code also enables, for example, the velocity of the laser to be adjusted to see how changes in this type of control input can impact the part (both in terms of the part geometry and stress profiles) and to suggest strategies for detecting when attacks that might impact these inputs occur. We discuss how different mesh sizes that are not necessarily at the level of mesh independence impact what can be learned from cyberattack studies, in particular in the case that major changes to the dynamics of the system occur (e.g., if the laser moves at a different speed), to better understand the computational resources required for a full analysis of cybersecurity concerns for additive manufacturing simulation.

REFERENCES:

[1] J. Plocher, and P. Ajit. "Review on design and structural optimisation in additive manufacturing: Towards next-generation lightweight structures." Materials & Design 183 (2019): 108164.

[2] J. Chang, J. He, M. Mao, W. Zhou, Q. Lei, X. Li, D. Li, C.K. Chua, and X. Zhao. "Advanced material strategies for next-generation additive manufacturing." Materials 11.1 (2018): 166.

[3] H.M. Khan, Y. Karabulut, O. Kitay, Y. Kaynak , and I. S. Jawahir. "Influence of the post-processing operations on surface integrity of metal components produced by laser powder bed fusion additive manufacturing: a review." Machining Science and Technology 25.1 (2020): 118-176.

[4] N. Ahmed, I. Barsoum, and G. Haidemenopoulos. "Process parameter selection and optimization of laser powder bed fusion for 316L stainless steel: A review." Journal of Manufacturing Processes 75 (2022): 415-434.

[5] M. Mani, B.M. Lane, M.A. Donmez, S.C. Feng, and S.P. Moylan. "A review on measurement science needs for real-time control of additive manufacturing metal powder bed fusion processes." International Journal of Production Research 55.5 (2017): 1400-1418.

[6] S. Y. Yu, A. V., Malawade, S. R. Chhetri, and M. A. Al Faruque. "Sabotage attack detection for additive manufacturing systems." IEEE Access 8 (2020): 27218-27231.

[7] K. Nieman, H. Oyama, M. Wegener, and H. Durand. "Predict the Impact of Cyberattacks on Control Systems." Chemical Engineering Progress. Sept. 2020.

[8] K. Nieman, D. Messina, M. Wegener, and H. Durand. "Cybersecurity and Dynamic Operation in Practice: Equipment Impacts and Safety Guarantees." Journal of Loss Prevention in the Process Industries. Submitted Feb. 15, 2021.

[9] K. Nieman, Leonard, A. F., K. Tyrrell, D. Messina, R. Lopez, H. Durand. “Challenges and Opportunities for Next-Generation Manufacturing in Space.” International Federation of Automated Control (IFAC) Symposium on Dynamics and Control of Process Systems (DYCOPS), In Press 2022.

[10] A. Hussein, L. Hao, C. Yan, and R. Everson. "Finite element simulation of the temperature and stress fields in single layers built without-support in selective laser melting." Materials & Design (1980-2015) 52 (2013): 638-647.

[11] K. Zeng, D. Pal, H. Gong, N. Patil, and B. Stucker. "Comparison of 3DSIM thermal modelling of selective laser melting using new dynamic meshing method to ANSYS." Materials Science and Technology 31.8 (2015): 945-956.

[12] J. Goldak, A. Chakravarti, and M. Bibby. "A new finite element model for welding heat sources." Metallurgical transactions B 15.2 (1984): 299-305.

[13] D. Deng and K. Shoichi. "Numerical simulation of residual stresses induced by laser beam welding in a SUS316 stainless steel pipe with considering initial residual stress influences." Nuclear Engineering and Design 240.4 (2010): 688-696.